Logistics Systems for Sustainable Cities Proceedings of the 3 r d International Conference on City Logistics (Madeira, Portugal, 25-27 June, 2003)
Related Elsevier books BREWER, BUTTON & HENSHER (eds.) Supply-Chain Management GIFFORD
Flexible Urban Transportation
HENSHER & BUTTON (eds.) ROOT (ed.)
Handbook of Logistics and
Handbook of Transport Modelling
Delivering Sustainable Transport
TANIGUCHI et al (eds) City Logistics: Network Modelling and Intelligent Transportation Systems
Related Journals Cities Editor: A. Kirby Transport Policy Editors: M. Ben-Akiva, Y. Hayashi and J. Preston Transportation Research Part B: Methodological Editor: F. Mannering Transportation Research Part E: Logistics and Transportation Review Editor: W. Talley For full details of all transportation titles by Elsevier please go to: www.elseviersocialsciences.com/transportation
Logistics Systems for Sustainable Cities Proceedings of the 3 r d International Conference on City Logistics (Madeira, Portugal, 25-27 June, 2003)
EDITED BY Eiichi Taniguchi Kyoto University, Japan Russell G. Thompson The University of Melbourne, Australia
2004
ELSEVIER Amsterdam - Boston - Heidelberg - London - New York - Oxford Paris - San Diego - San Francisco - Singapore - Sydney - Tokyo
ELSEVIER B.V. Sara Burgerhartstraat 25 P.O.Box 211, 1000 AE Amsterdam, The Netherlands
ELSEVIER Inc. 525 B Street, Suite 1900 San Diego, CA 92101-4495 USA
ELSEVIER Ltd The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB UK
ELSEVIER Ltd 84 Theobalds Road London WC1X 8RR UK
© 2004 Elsevier Ltd. All rights reserved. This work is protected under copyright by Elsevier Ltd, and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier's Rights Department in Oxford, UK: phone (+44) 1865 843830, fax (+44) 1865 853333, e-mail:
[email protected]. Requests may also be completed on-line via the Elsevier homepage (http://www.elsevier.com/locate/permissions). In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (+1) (978) 7508400, fax: (+1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (+44) 20 7631 5555; fax: (+44) 20 7631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of the Publisher is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier's Rights Department, at the fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made.
First edition 2004 Library of Congress Cataloging in Publication Data A catalog record is available from the Library of Congress. British Library Cataloguing in Publication Data A catalogue record is available from the British Library.
ISBN:
0-08-044260-9
@ The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Printed in The Netherlands.
V
PREFACE
Following the First and Second International Conferences on City Logistics that took place on 12th _ 14th July 1999 in Cairns, Australia, and in Okinawa, Japan on 27th - 29th June 2001, the Institute for City Logistics organised the Third International Conference on City Logistics in Madeira, Portugal on 25th - 27th June 2003. Urban freight transport has become an important issue in urban planning. There are many challenges and problems relating to increasing levels of traffic congestion, environmental impacts and energy conservation. In addition, freight carriers are expected to provide higher levels of service with lower costs. To address these complicated and difficult problems, numerous city logistics schemes have been proposed and implemented in several cities, including: co-operative freight transport systems, advanced information systems, public freight terminals and the regulation of load factors. City logistics schemes are relatively new concepts that are aimed increasing the efficiency of urban freight transport systems as well reducing traffic congestion and impacts on the environment. However, new modelling, evaluation and planning techniques are required to conduct in-depth investigations before city logistics schemes can be effectively deployed. This proceedings book includes recent developments in the modelling, evaluation and planning of city logistics schemes. Since city logistics schemes have already been implemented in several cities, a review of the performance of these schemes was presented and discussed. As well, an overview of the visions for city logistics and public private partnerships for city logistics was given. Recent developments in ICT (Information Communication Technology) and ITS (Intelligent Transport Systems) allows the efficiency of freight transport systems to be improved. ICT and ITS applications can integrate components for more efficient urban freight transport by private companies with transport policies oriented towards better urban environments promoted by the public sector. Therefore, ICT and ITS have good potential to promote public private partnerships for solving urban freight problems. We believe that this proceedings book covers wide range of important features of city logistics. It will help researchers, students and administrators to understand the current status of urban freight transport issues, models, evaluation methods and planning. We hope that the ideas and perspectives contained in this book will encourage people to research and implement schemes for creating more efficient and environmentally friendly logistics systems for sustainable cities.
vi
Preface
The Institute for City Logistics (http://www.citylogistics.org) has been active in undertaking research and development, organising conferences, workshops and short courses as well as publishing books in the area of city logistics. The Institute provides a platform for promoting exchanging knowledge, applying the new ideas and methods in modelling, evaluating and planning city logistics schemes. The Fourth International Conference on City Logistics will be organised by the Institute in 2005. We would like to express our heartiest appreciation to all the authors of papers submitted to the conference for their contributions and to the members of organising committee for their help in organising the conference.
Eiichi Taniguchi Russell G Thompson October 2003
THE ORGANISING COMMITTEE FOR 3 RD INTERNATIONAL CONFERENCE ON CITY LOGISTICS (MADEIRA, PORTUGAL, 25-27 JUNE 2003) Chair person Eiichi Taniguchi
Kyoto University, Japan
Russell G. Thompson
The University of Melbourne, Australia
Michael Browne
The University of Westminster, UK
Toshinori Nemoto
Hitotsubashi University, Japan
Tadashi Yamada
Hiroshima University, Japan
Ron van Duin
Delft University of Technology, The Netherlands
Johan GS.N. Visser
Delft University of Technology, The Netherlands
Kazuya Kawamura
University of Illinois, USA
Jose Holguin-Veras
Rensslaer Polytechnic Institute, USA
Dieter Wild
PTV, Germany
This page is intentionally left blank
CONTRIBUTORS Julian Allen University of Westminster, UK Louis Alligier Laboratoire d'Economie des Transports, France Christian Ambrosini Laboratoire d'Economie des Transports, France Stephen Anderson University of Westminster, UK Yasuo Asakura Kobe University, Japan Mem Baybars Transport for London (TfL) Street Management, UK Saurav Dev Bhatta University of Illinois, USA Daniel Bollo Inrets, Arcueil, France Michael Browne University of Westminster, UK Jose Mexia Crespo de Carvalho ISCTE - University of Lisbon, Portugal Georgina Christodoulou University of Westminster, UK Pablo Cortes University of Seville, Spain Alvaro Costa Universidade do Porto, Portugal Wanda Debauche Belgian Road Research Centre, Belgium J.H.R. van Duin Delft University of Technology, the Netherlands Gaetano Fusco Universita di Roma "La Sapienza"', Italy Simone Gragnani Federtrasporto, Rome, Italy Kim Hassall The University of Melbourne, Australia Eiji Hato Ehime University, Japan Makoto Hayano Docon Co. Ltd., Japan Katsuhiko Hayashi University of Marketing and Distribution Sciences, Japan Fred J.P.Heuer OECD Programme of Research on Road Transport and Intermodal Linkages Working Group on Urban Freight Logistics, Chairman, the Netherlands Rensselaer Polytechnic Institute, USA Jose Holguin-Veras Docon Co. Ltd., Japan Tatsuhide Ito Delft University of Technology, the Netherlands Milan Janic Henrik Enslev Jensen City of Copenhagen, Denmark University of Westminster, UK Peter Jones Osaka City, Japan Yasushi Kakimoto University of Illinois, USA Kazuya Kawamura Kilsby Australia, Australia David Kilsby Soeren Kjaersgaard City of Copenhagen, Denmark Delft University of Technology, the Netherlands J.C. Kneyber
x
Contributors
Uwe Kohler Oliver Kunze Juan Larrafieta Leorey Marquez Sandra Melo Kazuhiro Mori Jesus Munuzuri Toshiyuki Naito Toshinori Nemoto Luis Onieva Daniele Patier Raluca Raicu Serban Raicu Joan C. Rijsenbrij Jean-Louis Routhier Martin Ruesch Anusha Seetharaman Erwan Segalou Nariida Smith Marielle Stumm Eiichi Taniguchi Graham Tanner Luigi Tatarelli Mike Taylor Russell G. Thompson Gaetano Valenti Maria Pia Valentini Johan Visser Jaap VIeugel Tadayuki Wada Tony Whiteing Tadashi Yamada Yuji Yano Ryuichi Yoshimoto Yohei Yoshimura Rocco Zito
University of Kassel, Germany PTV AG Karlsruhe, Germany University of Seville, Spain CSIRO Australia, Australia Universidade do Porto, Portugal Hiroshima Institute of Technology, Japan University of Seville, Spain Docon Co. Ltd., Japan Hitotsubashi University, Japan University of Seville, Spain Laboratoire d'Economie des Transports, France University of South Australia, Australia Polytechnic University of Bucharest, Rumania Delft University of Technology, the Netherlands Laboratoire d'Economie des Transports, France Rapp Trans Ltd., Switzerland Cambridge Systematics, Inc./Volpe Center, USA Laboratoire d'Economie des Transports, France CSIRO Australia, Australia Inrets Arcueil, France Kyoto University, Japan University of Westminster, UK Universita di "Roma Tre", Italy University of South Australia, Australia The University of Melbourne, Australia ENEA, Rome, Italy ENEA- Ene/Tec, Centro Ricerche Casaccia - Roma, Italy Ministry of Economic Affairs, the Netherlands Delft University of Technology, the Netherlands Hokkaido Regional Development Bureau, Japan University of Huddersfield, UK Hiroshima University, Japan Ryutsu Keizai University, Japan Systems Research and Development Institute of Japan, Japan Hiroshima University, Japan University of South Australia, Australia
CONTENTS Preface 1 Visions for city logistics E. Taniguchi, R. G. Thompson and T. Yamada 2
Urban freight movements and public-private partnerships M. Browne, T. Nemoto, J. Visser and T. Whiteing
3 Transport demand, transport and traffic flow - Key elements of city logistics R. Raicu and S. Raicu 4
5
6
v 1
17
37
Assessment of the relationship between vehicle type mix and the benefit of freight proj ects K. Kawamura, A. Seetharaman and S. D. Bhatta
53
Estimation of an origin-destination matrix for urban freight transport. Application to the city of Seville J. Muñuzuri, J. Larrañeta, L. Onieva and P. Cortés
67
A new interactive approach on route planning with tight delivery time windows O. Kunze
83
7
Intelligent vehicle routing and scheduling R. G. Thompson
8
Road network reliability analysis using vehicle routing and scheduling procedures T. Yamada, Y. Yoshimura and K. Mori
111
On the estimation of the maximum efficiency of the trucking industry: Implications for city logistics J. Holguín-Veras
123
9
97
10 Modelling effects of e-commerce on urban freight transport E. Taniguchi and Y. Kakimoto
135
11 Last-mile, a procedure to set-up an optimized delivery scheme G. Fusco, L. Tatarelli and M. P. Valentini
147
xii
Contents
12 Towards a matching system for the auction of transport orders R. van Duin andJ.C. Kneyber
163
13 Systems theory, complexity and supply organizational models to enrich city logistics: An approach J. M. C. de Carvalho
179
14 Assessing impacts of greenhouse gas abatement measures on urban freight L. Marquez, N. Smith, D. Kilsby, M. Taylor and R. Zito,
191
15 The environmental assessment of urban goods movement E. Segalou, C. Ambrosini andJ.-L. Routhier
207
16 Route choice and the impact of'logistic routes' J. Vleugel and M. Janic
221
17 Empirical analysis on hazardous material transportation using road traffic census and accident data T. Ito, M. Hayano, T. Naito, Y. Asakura, E. Hato and T. Wada 18 Analysing the potential impacts of sustainable distribution measures in UK urban areas J. Allen, M. Browne, G. Tanner, S. Anderson, G. Christodoulou and P. Jones
235
251
19 Future city logistics in Japan from the shippers' and carriers' view - Prospects and recent measures to develop them K. Hayashi and Y. Yano
263
20
City logistics in Italy: A national project S. Gragnani, G. Valenti and M. P. Valentini
279
21 Developments in urban distribution in London M. Baybars and M. Browne
295
22 An experimental cooperative parcel pick-up system using the Internet in the central business district in Tokyo T. Nemoto 23 New ideas for the city-logistics project in Kassel U. Köhler
309
321
Contents xiii
24 A study on the setting up of lorry-dedicated routes in the Brussels Capital Region W. Debauche 25 New concepts for city logistics J. C. Rijsenbrij 26
27
Urban rail and intermodal freight strategies in the Zurich area: A case study from Switzerland M. Ruesch On-line retailing in France: Current and future effects on city logistics D. Patier and L. Alligier
28 Dispelling the e-commerce and urban transport environmental doomsday forecasts: A counter intuitive Australian case study - The postal transport network restructure, 1995 to 2000 K. Hassall 29 E-commerce and end delivery issues M. Stumm and D. Bollo 30 Web-based transport exchange systems in Japan and its implication to traffic volume R. Yoshimoto 31 Summary of the OECD report 'Delivering the Goods-21st Century Challenges to Urban Goods Transport' OECD Programme of Research on Road Transport and Intermodal Linkages Working Group on Urban Freight Logistics
333
349
365
381
397
405
421
431
32 Sustainable city logistic solutions S. Kjaersgaard and H. E. Jensen
441
33 The E+ Transport environmental operator classification system K. Hassall
449
34 Relationships between goods distribution and public transport in urban areas - the case of a hypermarket in Porto A. Costa and S. Melo
459
This page is intentionally left blank
1
VISIONS FOR CITY LOGISTICS Eiichi Taniguchi, Kyoto University, Japan Russell G. Thompson, The University of Melbourne, Australia Tadashi Yamada, Hiroshima University, Japan
ABSTRACT This paper presents visions for city logistics that are required to set targets of the activities that can be achieved using city logistics schemes. Our visions for city logistics consider three pillars that are guiding principles: (a) Mobility, (b) Sustainability and (c) Liveability. These three pillars are supported by goals that brace the structure of the visions, comprising: (a) Global competitiveness, (b) Efficiency, (c) Environmental friendliness, (d) Congestion alleviation, (e) Security, (f) Safety, (g) Energy conservation and (h) Labour force. This paper discusses various features associated with urban freight transport issues to create mobile, sustainable and liveable cities. It concludes that there are a number of promising schemes that have the potential to fully realise the visions of city logistics, including: (i) Establishing effective partnerships between key stakeholder groups, (ii) Implementing information and communication technology and intelligent transport systems, (iii) Promoting corporate responsibility, (iv) Incorporating urban freight transport as an integral component of urban planning.
INTRODUCTION Why are visions necessary? This paper will present visions for city logistics. Why do we need visions? We have already given the definition of city logistics in the 2nd International Conference on City Logistics in Okinawa, Japan as: "City Logistics is the process for totally optimising the logistics and transport activities by private companies with support of advanced information systems in urban areas considering the traffic environment, the traffic congestion, the traffic safety and the energy savings within the
2
Logistics systems for sustainable cities
framework of a market economy" (Taniguchi et ah, 2001) This statement gives us a conceptual idea of what is city logistics. However, in order to establish efficient and environmentally friendly urban logistics systems through the process of city logistics, we need visions for city logistics. First of all, it is necessary to set targets of the activities that can be achieved using city logistics. In this context we would like to consider three pillars as shown in Figure 1: (a) Mobility (b) Sustainability (c) Liveability Mobility is a basic requirement for transporting goods within as well as into and from urban areas. Reliable road, rail and other modal network are essential in terms of connectivity and travel times. Providing enough road network capacity and alleviating traffic congestion is always important in the agenda of urban traffic management. In particular this is vital for urban freight transport, since many of freight carriers have to meet severe time windows set by customers within the framework of Just-In-Time transport systems. Sustainability has become more important, since people are concerned about environmental issues including air pollution, noise, vibration and visual intrusion. Large freight vehicles are often the source of these negative environmental effects. Therefore, minimising the negative impacts on the environment by trucks is an important issue to be addressed when managing urban freight transport systems. As well, minimising energy consumption is required to ensure a sustainable city. Liveability should be taken into account when planning urban logistics systems. Residents in urban areas enjoy the benefits of buying wide variety of commodities based on urban delivery systems to retail shops or even directly to homes. But they are also concerned about traffic safety and environment in community, which may be threatened by heavy commercial vehicles travelling within and near residential areas. Therefore, the visions for city logistics is to create a mobile, sustainable and liveable city by supplying necessary goods for activities and collecting goods that are produced in the city as well as minimising negative impacts on the environment, safety and energy consumption.
Visions for city logistics
3
As Figure 1 indicates, mobility, sustainability and liveability are three pillars of the visions for city logistics. They are supported by goals that brace the structure of the visions, comprising: (a) Global competitiveness (b) Efficiency (c) Environmental friendliness (d) Congestion alleviation (e) Security (f) Safety (g) Energy conservation (h) Labour force The pillars are the guiding principles of city logistics. They represent the philosophy of city logistics. The pillars provide the strategic basis for planning and managing urban goods movement systems. Goods movement has a strong influence on the sustainability, mobility and liveability within urban areas. City logistics embraces these planning principles and strives to enhance them.
4
Logistics systems for sustainable cities
The braces or cross bars provide the goals or directions for city logistics to address common issues. They are derived from societal values. Each goal or cross-bar will relate to one or more key stakeholder. This illustrates the complexity of city logistics where there are often multiple goals that maybe seen to be conflicting. The goals must be considered in city logistics to achieve more mobile, sustainable and liveable cities. Urban freight problems are varied and complex since they are associated with economic, financial, environmental and social aspects. Creative and innovative schemes are required to ensure that goods movement in cities enhances accessibility and the wellbeing of residents as well as not jeopardising the potential for future generations to enjoy a similar or improved quality of life. City logistics schemes should encapsulate the visions by enhancing the principles and achieving the goals of key stakeholders. The visions should inspire researchers and practitioners to develop and implement solutions since it provides a focus for guiding the development of urban logistics systems to make cities better places to live, do business and socialise. ICT, ITS and city logistics hi 20th century there was considerable discussion that focused on the trade-off relationship between efficiency and environmental friendliness of urban freight transport systems. That is, if freight carriers try to establish more efficient urban freight transport systems, they will generate more negative impacts on the environment. In contrast, if they try to use environmentally friendly urban delivery systems, they have to pay extra costs for lowering hazardous gas emissions and noise (e.g. Cooper, 1991). However, in 21st century we are in a good position of being able to use Information Communication Technology (ICT) and Intelligent Transport Systems (ITS) for overcoming the trade-off relationship between efficiency and environmental friendliness. ICT and ITS enable efficient and environmentally friendly urban delivery systems to be developed. An example of this is ITS based probabilistic or dynamic vehicle routing and scheduling planning with time windows. Taniguchi et al. (2001) showed that this approach allows freight carriers to reduce their costs of operating pickup/delivery trucks as well as reducing CO2 emissions. In the future, advanced information systems will be more widely applied in logistics industry allowing both efficient and environmentally friendly delivery systems to be realised. City planning and city logistics Freight transport should be explicitly considered in urban planning. However, many of cities have not yet established a master plan for urban freight transport. This is attributed to the lack of appropriate personnel, knowledge and data for urban freight transport (OECD, 2003). Not surprisingly many of cities have no section that is fully devoted to addressing urban freight issues. In addition, city planning normally considers long term issues over a period of several years, while logistics planning horizon of private companies is more short term often over a few months. This mismatch of the length of planning periods often makes it difficult for freight issues to be effectively addressed by both city authorities and private companies. As urban freight transport is very important for the sustainable development in urban areas, cities
Visions for city logistics
5
should pay more attention for including it in their city planning. Each city needs to draw comprehensive visions for urban freight issues. To do this, city planners should be aware of appropriate knowledge and know-how on urban freight issues and collect necessary data. Land use planning and city logistics A common problem in city planning involves the inadequate location of logistics facilities. Freight carriers have difficulty in finding appropriate land for logistics terminals in urban areas because of zoning regulations and land prices. If they built their logistics terminals far from an expressway interchange and if the area between the interchange and the logistics terminals is developed as a residential area, it generates environmental and safety problems due to large trucks travelling on the roads through the residential area. This type of problem can be solved by incorporating freight transport plans within land use planning processes.
Units of urban freight transport planning Another problem related to urban freight transport planning is associated with the unit of planning. Basically each individual authority makes their own urban traffic plans that are typically orientated towards passenger cars. But the unit of city is in reality often too small for freight transport planning, since freight transport in one city area is only a part of supply chain that includes line haul transport between cities or even between countries. Therefore, urban freight transport planning in a city should be harmonised with other adjacent cities. Several cities in a wide area or in a corridor should preferably discuss their ideas of urban freight transport planning to make a common plan for freight transport. Such an institutional framework in a wider area for urban freight transport planning is essential for sustainable development of urban
Subsidies and additional charges from the public Urban freight transport is a private sector activity undertaken by companies. In principle neither subsidies nor additional charges should be required to be made by the public sector to private companies. However, freight transport has a large influence on the economic development of cities as well as social and environment systems. In some cases the central government and city authorities assist freight carriers and shippers financially by providing subsidies or low interest loans. In other cases they impose additional charges to improve the environment. This public commitment is based on the following reasons: (a) It is difficult to internalise the external benefits by new investment (e.g. introducing low emission trucks by freight carriers) (b) It is hard to internalise the external diseconomy by operating freight vehicles (e.g. congestion charging) (c) A huge initial investment maybe required to start a large freight transport project (e.g. building a large scale logistics terminal for co-operative delivery systems by small and medium size enterprises) Subsidies are sometimes provided to freight carriers who use low emission trucks (e.g. compressed natural gas vehicles) in urban areas. Charging vehicles coming into central areas has
6
Logistics systems for sustainable cities
been considered in many large cities and was implemented in London in 2003. Subsidies are commonly provided for encouraging small and medium size enterprises to participate in innovative logistics projects, such as advanced information systems and co-operative delivery centres.
MOBILITY It is acknowledged that freight transport is essential to the sustainable development of cities. However, especially in urban transport planning, its importance is often overlooked (e.g. Button and Pearman, 1981; Ogden, 1992). Urban traffic problems have been highlighted in the political agenda for some time, and passenger movements have been considered to play a much more important role than freight. Mobility is defined as ease of movement (Jones, 1981; Levine and Grab, 2002). Indeed the enhancement of mobility for personal trips is a matter of great importance, but in fact, local economic growth and higher quality of daily life depends in part on efficiently transporting goods in urban areas. It is therefore crucial to enhance the mobility for freight vehicles. Characteristics of urban freight transport Freight transport within urban areas has different characteristics from inter-city long haulage. The main attributes of urban freight transport can be summarised as: (a) (b) (c) (d)
Frequent delivery of smaller quantities Low utilisation of the capacity of trucks Time windows On-street parking
Frequent delivery of smaller quantities- A notable trend in urban freight transport is the control of deliveries within the framework of supply chain management (SCM). The JIT (just-in-time) concept was originally formulated to aid production processes in the automobile industry, being introduced by Toyota Motors for its assembly line operation (Giannopoulos and McDonald, 1997). JIT delivery concepts have now been implemented in all industrial sectors, since it allows better inventory control and cost reduction. Consequently, deliveries have become more frequent, with multiple trips often being made by vehicles, and the quantities delivered correspondingly smaller. Low utilisation of the capacity of trucks- The OECD working group on urban goods transport (OECD, 2003) estimated that around 30 percent of vehicles carry loads of 25 percent below capacity, and 50 percent loads of more than 50 percent below capacity. Giannopoulos and McDonald (1997) also identified that the load factors of small trucks performing urban deliveries in urban areas in Japan is around 20 percent, whilst that of ordinary trucks is around 60 percent. Reasons for this included that a large number of private cars are used for urban freight transport on a daily basis as well as that the frequent delivery of smaller quantities has been undertaken to provide customers with a higher quality of service. On-street parking- The lack of parking space for loading/unloading is clearly noticeable in city centres. It has been estimated in various surveys undertaken in European cities that, where
Visions for city logistics
7
controls are not sufficiently strict, the percentage of illegal parking can reach a level of 65-70 percent (OECD, 2003). Odani and Tsuji (2001) also reported that nearly 95 percent of truck drivers usually find parking spaces on the road for goods delivery in the city centre of Kobe. Time windows. Freight carriers are often requested to arrive at customers within a specified time period, since the reliability of goods delivery has become important in recent urban freight transport systems (e.g. JIT delivery systems). Strict time windows have led to smaller loads of goods being transported more frequently. A recent survey in Osaka and Kobe in Japan found that freight carriers were requested to operate with designated arrival times or time windows for 52 percent of goods delivered and for 45 percent of goods collected in terms of weight (Taniguchi et al, 2001) Urban road congestion Urban transport systems are characterised by dense traffic networks. Road congestion is becoming a major problem in developing cities as well as in developed cities. Travel speeds within urban areas of major cities are typically in the range of 10-30 km/h. Freight vehicles comprise 20-30 percent of the total traffic volume in most developed cities (Kenworthy and Laube, 1999). The proportion of freight vehicles on urban motorways is on average 6-15 percent, while in city centres in peak hours this can reach 20-22 percent (OECD, 2003). Freight vehicles tend to have greater trip lengths than cars. In Tokyo Metropolitan Region, freight vehicles occupy approximately 40 percent of the total length of vehicle trips, in terms of vehicle-kilometres (Committee for ICT based sustainable urban distribution systems, 2003). This means that trip length per vehicle of freight vehicles is larger than that of passenger cars. Freight movements using road-based vehicles are therefore, an important source of road congestion within urban areas. The OECD working group on urban goods transport (OECD, 2003) illustrates the major causes of urban road congestion as follows: (a) (b) (c) (d) (e)
Inadequate infrastructure Too many vehicles at the same time Time windows Opening hours of shops Interaction between commercial vehicles and private cars
Freight carriers often face difficulty in operating their vehicles on urban roads due to traffic congestion. This has led to the inefficient use of trucks, where smaller loads are being transported. Trucks are obliged to wait near the location of customers when they arrive earlier than the time specified by customers. Accessibility and reliability Freight traffic now has to satisfy the social needs as well as being economically efficient. Freight vehicles in urban areas generates high societal costs by producing pollution and noise. Although trucks account for only 10 percent of all transport operations in urban areas, they produce over 40 percent of the pollution and noise caused by local traffic (OECD, 2003). This is largely due to
8
Logistics systems for sustainable cities
the frequent deliveries of smaller quantities as well as the higher release of contaminants from engines. Parking without switching off the engine also contributes to such environmental problems. Nuisance problems other than air pollution and noise are often caused by urban freight traffic, including road crashes, physical hindrance and vibration. Road crashes deteriorate the reliability and accessibility of freight transport, since they cause traffic congestion and sometimes make road links impassable. Accessibility is a function of mobility, which can be defined as the ease of reaching destinations (Jones, 1981; Levine and Grab, 2002). Reliability in transport systems, describes the possibility of successfully travelling from one place to another (Billington and Allan, 1992; Wakabayashi and Iida, 1992). Decreased accessibility and reliability leads to the reduction in levels of service and efficiency of freight transport. Freight vehicles generate urban traffic problems, and consequently these problems produce inefficient freight movement within urban areas.
SUSTAINABILITY Overview Environmental issues and energy conservation need to be taken into account when applying city logistics measures for sustainable cities. Noise, air pollution (CO, NOx, SOx, CO2, PM, HC), vibration and visual intrusion are major environmental issues related to freight vehicles. Minimising the fuel consumption in operating commercial vehicles for urban goods delivery is also important. The harmonisation of efficiency and environmental friendliness and energy conservation is essential for ensuring sustainable development in urban areas. Normally there are four stakeholders in urban freight issues; (a) shippers, (b) freight carriers, (c) residents and (d) administrators. Each of these key stakeholders has their own specific objectives and tends to behave in different manner (Thompson and Taniguchi, 2001). Therefore, city logistics measures should incorporate these features. Innovative solutions There are several innovative city logistics schemes for enhancing the sustainability of urban areas including: (a) Use of ICT and ITS (b) Environmental appraisal (c) Co-operation Use of ICT and ITS The development and deployment of ICT (Information Communication Technology) and ITS (Intelligent Transport Systems) can allow shippers and freight carriers to reduce logistics costs as well as improving the environment by reducing hazardous gas emissions. Recently traffic information on travel times is available using ITS. For example, VICS (Vehicle Information
Visions for city logistics
9
Communication Systems) in Japan provides real time travel times to drivers. Already 6.58 million vehicles (about 8.5% of all vehicles) are equipped with VICS in Japan in March 2003. Once these real time travel times data are stored in a computer, the distribution of travel times can be determined as shown in Figures 2 and 3. Both historical and real time information on travel times is useful to upgrade the conventional vehicle routing and scheduling planning. Probabilistic vehicle routing and scheduling uses historical travel times data and considers the uncertainty of link travel times on the road network (Taniguchi and Thompson, 2002). As well, dynamic vehicle routing and scheduling uses real time travel times (Taniguchi et al, 2001). These methods are effective for decreasing the costs for freight carriers by optimising the visiting order and route to visit customers by trucks. Consequently these planning methods lead to the reduction of total running times of trucks and the number of trucks used.
Figure 2 Example of change in travel times on a link for a national highway in Japan measured by VICS for one year
10 Logistics systems for sustainable cities
Figure 3 Example of distribution of travel times on a link of a national highway in Japan by VICS for one year Logistical matching systems via the Internet present a good opportunity to connect supply and demand for goods transport. Shippers can offer a job for carrying goods and freight carriers can offer vacant space on their trucks. Matching systems allow freight carriers to increase the load factor of their trucks as well as shippers to quickly find an appropriate freight carrier to transport their goods at a lower price and better service. Therefore, matching systems have great potential to improve the efficiency of urban freight transport and reduce the vacant truck trips, which will lead to lower negative impacts on the environment. Although the use of matching systems is limited to the spot market at the moment, this type of e-commerce of logistics transactions will be very helpful for sustainable development of urban areas in the future. Environmental appraisal There are attempts to encourage and evaluate the efforts of companies to improve the environment. Some shippers and freight carriers are concerned about global and local environment issues. Companies who have a certificate of ISO (International Organisation for Standardisation) 14000 series, use environmental management systems where "plan, do, check and action" process have been identified, hi particular, ISO 14031 presents the process for collecting data and information for evaluating environmental performance. Another example is an agreement between a company and WWF (World Wide Fund for nature) on reducing CO2 emissions. A company who signs this agreement sets a target of reducing CO2 emissions and publishes the results of performance, for example, by introducing low emission vehicles. Although these are voluntary based programs an increasing number of companies are involved in the process for improving the environment. The main reasons for companies participating in such schemes are to create a good image to consumers as well as the potential of gaining accessing to green funds. Such schemes provide loans at low interest to companies who achieve good environmental performance.
Visions for city logistics
11
Co-operation Co-operation among competitive shippers and freight carriers is essential for city logistics. The concept of "co-opetition" (Brandenburger and Nalebuff, 1996) that is combination of co-operation and competition, gives a good mindset for creating co-operative freight transport systems. For example, 11 department stores in Osaka area, Japan started co-operative home delivery systems to customers in 1989 (Taniguchi and Nemoto, 2002). In this project, after discussing the problems relating home delivery of commodities, the presidents of department stores made a decision not to compete in the area of home deliveries, while they are still competitive in the area of sales. Therefore, this is a system of competition in sales but co-operation in delivery. The system was very successfully applied and produced good benefits by reducing the distance travelled by pickup/delivery trucks and labour hours.
LlVEABILITY What is Liveability? Liveability relates to the quality of life experienced. It is an ambiguous concept that incorporates peoples satisfaction or gratification with their life and external conditions (The British Library, 1994:x) Liveability involves an assessment of the conditions that are experienced and interpreted within an individuals life sphere. It incorporates not only the physical characteristics of place but social attributes of the environment, hi a broader sense liveability involves evaluating a cities environment. Liveability relates to the degree of comfort encountered as well as image, including many intangibles such as attractiveness, safety, peacefulness and charm.
Measuring liveability Hardship ratings regularly produced by private sector organisations for expatriates are used by governments and international companies to determine allowances that reflect differences in the quality of life for personnel living abroad. Rankings based on desirability criteria are produced using multi-criteria models combining criteria and weightings. A number of factors are typically included such as the political, social, environmental, financial and economic conditions as well as personal safety, education, health care, housing, shopping and transport. Air quality and congestion levels are usually incorporated in liveability ratings. Air pollution, noise and crashes are the most obvious direct factors relating to goods movement influencing liveability in cities. There is a need to define what needs to be measured and how to measure it. City Logistics is based on the systems approach that involves defining problems and stating objectives (Taniguchi era/, 2001). Recent European initiatives have been orientated towards developing standardised monitoring programs (e.g. noise statistics) and presentation of results (e.g. noise maps). However, there are inherent weaknesses in using descriptive statistics in explaining social phenomenon.
12 Logistics systems for sustainable cities Road traffic is typically the most important source of noise pollution in urban areas. Noise statistics are often difficult to interpret and do not always link directly with annoyance, discomfort and anxiety. There can be a significant variation in individuals recognition, perception and tolerance of problems. A more meaningful set of freight related indicators that adequately reflect the quality of life in cities needs to be defined. People orientated approaches that account for conditions actually experienced should be developed. There are difficulties developing indicators that encapsulate the variation of impacts that are experienced within a city. Aggregating simple criteria based on location specific measurements tends to loose important information. Indirect, informal perception based indicators have potential to be cost effective in tracking progress and providing a snapshot of quality of life. For example, property values and turnover rates could be used to measure satisfaction with living in particular areas. Regular attitude surveys where people rate their satisfaction seem to provide richer information relating to liveability.
Goals To improve both the quality of life and economic growth for urban dwellers Cities are developed to exchange culture, goods, friendship, knowledge, emotional and spiritual support and minimise travel. Within cities, people work, rest and play. Due to the growth in urbanisation the quality of life for most people in the future will be determined by the quality of life in cities. However, urban life is a paradox, people congregate in cities in response to opportunities but this creates undesirable experiences. Liveability and economic growth are both vital for sustainable cities. Cities are more than centres of economic production and consumption, urban areas have a number of social, economic, cultural, physical, political and environmental aspects. Strategies for improving liveability within cities must recognise that the diverse dimensions of urban areas are closely linked and initiatives that address one will have an effect on others. Zero Tolerance Apart from the physical danger associated with truck travel, the noise, vibration, air pollution, dust and dirt generated interferes with urban living. Clean air and a good nights sleep are rights, not privileges for all urban dwellers. There should be no negative impacts of goods movement in cities of the future. Freight transport should occur by stealth. Goods movement within cities should be be undetectable by human senses. Develop freight systems that enhance values Future freight systems in cities need to enhance, not threaten values, including: (i) (ii)
Health Peace and quiet (sittability, sleepability and walkability)
Visions for city logistics (iii) (iv)
13
Safety Equity
Minimise the impacts of freight in cities Traffic noise is becoming an increasingly worrying irritant in most cities. Residents in many cities are exposed to high noise levels generated by freight vehicles. Excessive noise levels commonly are produced by engines, transmissions, exhausts, interaction between the tyres and the road pavement as well as load rattles. Stress, anxiety and other illness can be attributed to noise generated by freight vehicles. Night deliveries or spreading of the freight peak would have beneficial effects in terms of congestion however, it could cause detrimental effects for impacts such as noise pollution. Diesel fumes can cause serious health problems. Exposure by residents to exhaust gases and suspended particle matter is dangerous. More efforts should be made to reduce the amount of harmful gases produced from freight vehicles and limit the amount of freight vehicle movements in and near residential areas. Partnerships Contemporary planning is based on creating partnerships between key stakeholders. Governments cannot rely on prescriptive regulations but should create effective relationships for reducing the impacts of freight movement within cities. Effective collaboration between industry, government and knowledge institutes will ensure that the complex freight related problems in urban areas can be solved. Communities, government and industry need to work together to solve complex urban problems and raise the profile of urban freight. "In improving goods movement in Asian cities, a comprehensive approach is necessary which recognises that urban freight handling is first and foremost an inherently private sector activity composed of many closely inter-related elements and that freight movements are just as important as passenger movements". (Midgley:7, 1994) How can we civilise the truck? More effort should be made to make trucks more compatible with human activities in urban areas. Current vehicle standards generally do not realistically address the potential impacts of truck operations on communities. For example, vehicle standards often do not consider operational aspects such as braking and accelerating. Badly maintained or adjusted engines can be substantial polluters. Environmental management systems (EMS) can assist the private sector to take responsibility for their actions. EMS are based on implementing environmental quality control and monitoring systems (e.g. ISO 14000). EMS are a voluntary based approach that can be tailored for specific tasks by carriers. EMS for urban carriers have good potential to address a wide range of community concerns. A three tiered environmental operator performance rating scheme has recently been developed for urban distribution in Australia (Hassall, 2002). This 'green fleet' or E+ rating system reflects the effort an operator puts into environmental initiatives, e.g. cleaner engines and waste disposal policies. Such schemes will facilitate responsible environmental practices by operators in urban areas without regulation.
14 Logistics systems for sustainable cities Reduce vehicular demand Most regulations for addressing the impacts of trucks in cities are aimed at reducing the generation of impacts at the source (e.g. vehicle design standards). There are few comprehensive programs orientated towards reducing the number of vehicle movements. Recent initiatives in performance based standards, load factor controls and road pricing will provide incentives for operators to reduce vehicle movements. Co-operative based systems combined with public logistics terminals and advanced information technology have good potential to reduce vehicle movements. Such measures will increase the utilisation of trucks and stimulate the production of more goods in the areas where they are consumed.
Strategies for enhancing liveability The application of advanced technology such as intelligent transport systems offers potential for reducing the number of freight vehicles travelling in cities. Underground freight systems eliminate many of the effects of goods movement on liveability. However, they are very expensive. Existing underground facilities maybe able to be adapted to have a freight function. There are numerous initiatives that could be implemented to enhance liveability in urban areas (Taniguchi and Thompson, 2002). Low emission vehicles could be mandated. New information based technologies could be applied to develop more intelligent freight systems (Thompson, 2002). Non-motorised freight vehicles such as trolleys, carts and cycles can provide effective distribution between local centres and homes. Public transport systems including subways could be adapted to have a freight capacity.
Equity Night deliveries may be a solution to avoid congestion and minimise visual intrusion of trucks but there are serious social issues associated with evening-shift work. Disruption to family patterns and reduced enjoyment of daylight conditions are important issues. The poor have a limited choice of housing and tend to live in unattractive areas where the impacts from freight vehicles are highest, e.g. near busy roads and railway lines. The experiences of deprived groups need to be captured not only the aristocratic elite. Economic growth can threaten liveability within cities. "The challenge is to develop and implement policies that support not only the function of cities as engines of economic growth, but also their role as agents of social change" (UN, 2001:235)
CONCLUSIONS The three pillars of visions for city logistics, mobility, sustainability and liveability braced by values, such as safety, security and economic prosperity should inspire researchers and practitioners to develop and implement solutions for solving urban freight problems.
Visions for city logistics
15
Contemporary urban logistics systems provide a wide range of benefits for residents. However, significant negative impacts can arise. There are a number of promising schemes that have the potential to fully realise the visions of city logistics, including: (i) Establishing effective partnerships between key stakeholder groups (ii) Implementing information and communication technology and intelligent transport systems (iii) Promoting corporate responsibility (iv) Incorporating urban freight transport as an integral component of urban planning The challenge is discover innovative schemes and develop planning processes that will allow the visions of city logistics to be achieved.
REFERENCES Billington, R. and R.N. Allan (1992). Reliability Evaluation of Engineering Systems: Concepts and Techniques. Plenum Press, New York. The British Library (1994). Quality of life in cities - An overview and guide to the literature, London. Brandenburger, A. M. and B. J. Nalebuff (1996). Co-opetition, Doubleday, New York. Button, K. and A. Pearman (1981). The Economics of Urban Freight Transport- Macmillan, London. Cooper, J. (1991). Innovation in logistics: the impact on transport and the environment, In M. Rroon, R. Smit and J. van Ham (Eds.) Freight Transport and the Environment, Elsevier, pp.235-254. Committee for ICT based sustainable urban distribution systems (2003). Sustainable urban distribution systems. The Infrastructure Planning Committee, JSCE. (in Japanese) Giannopoulos, GA. and M. McDonald (1997). Developments in transport telematics applications in Japan; traffic management, freight and public transport. Transport Reviews, 17(1), 37-59. Hassall, K. (2002). Green rating incentives, Australian Transport News, July 26, 2002, 16. Jones, S.R. (1981). Accessibility measures: A literature review. TRRL Report 967, Transport and Road Research Laboratory, Crowthorne, Berkshire. Kenworthy, J. and F. Laube (1999). An International Sourcebook of Automobile Dependence in Cities, 1960-1990. University Press of Colorado, Boulder, Cololado. Levine, J. and Y. Grab (2002). Congestion pricing's conditional promise: promotion of accessibility or mobility? Transport Policy, 9, 179-188. Midgley, P., (1994). Urban transport in Asia - An operational agenda for the 1990's, World Bank Technical Paper Number 224. Odani, M and T. Tsuji (2001). An experiment to demonstrate the effectiveness of on-street parking facilities for delivery vehicles. City Logistics II (E. Taniguchi and R.G Thompson, eds.), 351-365, Institute for city logistics, Kyoto. OECD working group on urban freight logistics (2003). Delivering the goods 21st century challenges to urban goods transport, OECD. Ogden, K. (1992). Urban Goods Movement. Ashgate, Aldershot. Taniguchi, E., Thompson, R.G, Yamada, T. and R. van Duin. (2001). City Logistics: Network Modelling and Intelligent Transport Systems, Pergamon, Oxford. Taniguchi, E., R.G Thompson and T. Yamada (2001). Recent advances in modelling city logistics, In E. Taniguchi and R.G Thompson (Eds.) City Logistics II, Institute of Systems Science Research, Kyoto, pp.3-34.
16 Logistics systems for sustainable cities Taniguchi, E. and R.G Thompson (2002). Modelling city logistics, Transportation Research Record, Journal of the Transportation Research Board, No. 1790, pp.45-51. Taniguchi, E. and T. Nemoto (2002). Transport demand management for freight transport, In E. Taniguchi and R.G. Thompson (Eds.) Innovations in Freight Transport, WIT Press, Southampton, pp. 101-124. Taniguchi, E. and R.G Thompson (Eds.) (2002). Innovations in Freight Transport, WIT Press, Southampton. Thompson, R.G and E. Taniguchi, (2001). City Logistics and Freight Transport, in Handbook of Logistics and Supply Chain Management, Handbooks in Transport, Vol. 2, Elsevier, Oxford (A.M. Brewer, K. Button and D.A. Hensher, Eds.), 393-404. Thompson, R.G (2002). Intelligent Freight Systems for Sustainable Cities, Proc. 2nd Best Urban Freight Solutions (BESTUFS) Conference, Paris. UN (2001). Cities in a Globalizing World, Global Report on Human Settlements, United Nations Centre for Human Settlements (Habitat), Earthscan, London. Wakabayashi, H. and Y. Iida (1992). Upper and lower bounds of terminal reliability of road networks: an efficient method with Boolean algebra. Journal of Natural Disaster Science, 14 (1), 29-44.
2
URBAN FREIGHT MOVEMENTS AND PUBLICPRIVATE PARTNERSHIPS
Michael Browne, Transport Studies Group, University of Westminster, London, UK Toshinori Nemoto, Graduate School of Commerce and Management, Hitotsubashi University, Tokyo, Japan Johan Visser, Ministry of Economic Affairs, Den Haag, The Netherlands Tony Whiteing, Division of Transport and Logistics, University of Huddersfield, Huddersfleld, UK
ABSTRACT This paper reviews the increase in public-private partnerships (PPPs) in urban distribution in recent years. A discussion of various approaches to PPP is included, together with consideration of the forms that participation can take. The ways in which PPPs have been applied to urban distribution is considered, with detailed examination of several strategies and policies.
INTRODUCTION Public-private partnership (PPP) has become a popular concept in the public sector in many countries during the last decade. This has manifest itself in two forms: (i) particular projects in which the public and private sector have shared interests and objectives and where there is often an element of shared risk and reward, and (ii) initiatives between the public and private sector that involves information dissemination, communication, co-operation or joint working. The former (narrower) type of PPP helps to increase the amount of long-term private sector involvement in public sector businesses and projects that the public sector has responsibility for. The latter (broader) type of PPP is becoming more widespread and is favoured by government as it ensures stakeholder participation in policy decision-making, and thereby, it is hoped, results in greater success in the implementation of new initiatives.
18
Logistics systems for sustainable cities
Both approaches to PPP can be seen within the field of urban freight transport, especially the broader type. This is explained by the fact that urban freight comprises many different stakeholders with diverse interest (including retailers, wholesaler, carriers, warehousing, residents, shoppers and workers). The global movement of people, goods and information has further accelerated the extent of diversification, which makes our lives exciting, for example, by offering the consumer many choices. However, public decision-making has required more efforts to co-ordinate these activities to ensure that they function efficiently, while at the same time minimising the social and environmental impacts associated with them. In order to attempt to reach democratic decisions that will achieve these objectives, policy makers have been working closely with stakeholders on a range of urban freight issues. This paper discusses both types of PPP, and considers how PPPs have been applied to urban freight transport. First, we define the terms 'partnerships' and 'public-private partnerships'. Second, we consider the application of PPPs in the field of urban distribution, both in terms of narrow and broad PPPs. The range of stakeholders in urban distribution is discussed as this raises issues concerning PPP organisation and success. Strategies and measures relating to both the public and private sector are presented and the potential for collaboration discussed. Specific policy measures and strategies that involve PPPs are then examined in some detail using examples. These include urban transhipment centres, intermodal centres, alternative power vehicles, Freight Quality Partnerships, intelligent transport systems, Low Emission Zones and congestion charging.
PUBLIC-PRIVATE PARTNERSHIPS
Partnerships Partnership has become a very commonly used word in government planning and policy in the last few years. This is reflected in the first definition of the word taken from the Oxford English Dictionary shown below: "1. The fact or condition of being a partner; association or participation. Now esp. of relationships in industry and politics. 2.0. An association of two or more persons for the carrying on of a business, of which they share the expenses, profit, and loss. b. The persons collectively composing such a business association. 3. The rule or method for the calculation of a partner's share of gain or loss in proportion to his share of the capital or other determining conditions. " When used in relation to public-private sector initiatives it is sometimes used in a narrow sense relating to sharing expenses, profit, and loss, but it is often used in a broader sense to mean information dissemination, communication, co-operation or joint working by the public sector and other organisations and individuals. As Lowndes (2001) noted, "Partnership refers to a variety of arrangements with different purposes, time-scales, structures, operating procedures and members. A partnership may
Urban freight movements and public-private partnerships
19
simply be a means of 'getting people together' to begin a debate or share information, or it may be a policy-making forum, or even a contractually-based arrangement for service delivery." Lowndes (2001) put forward three reasons for the increased use of partnership by policymakers (see Box 1).
Public-private partnerships A great variety of meanings are attached to the phrase 'Public-Private Partnerships' (PPPs). Sometimes PPP is used in a very narrow way that is concerned with particular projects in which the public and private sector have shared interests and objectives and where there is often an element of shared risk and reward. These partnership schemes were introduced by certain governments in the 1990s for the purpose of increasing the amount of long-term private sector involvement in public sector businesses and projects that the public sector has responsibility for. The UK government's definition of this type of PPP is shown in Box 2.
20 Logistics systems for sustainable cities
Transport projects funded in the UK as part of PPPs include: • • • •
river crossings road constructions rail extensions tramway construction
Similar PPP arrangements have been used in other European countries: •
In France, PPPs have been used to supply community services and road building (Erlach, 2002)
•
In Austria, PPPs have provided communal infrastructure such as ASFINAG, the government-owned motorway development and management company, the Ebelsberg bypass in Linz, and the combined cargo traffic in Werndorf (Erlach, 2002) In Italy, the planned logistics intermodal terminal in Parma (Sardi, 2002).
•
The Canadian Council for Public-Private Partnerships has adopted the following definition in an attempt to clarify what is meant by the PPP concept: "[a] co-operative venture between the public and private sectors, built on the expertise of each partner, that best meets clearly defined public needs through the appropriate allocation of resources, risks and rewards. " However, the term PPP is also regularly used in a much broader sense to mean any initiative between the public and private sector that involves information dissemination, communication, co-operation or joint working joint. In the UK, for example, the Prime Minister, Tony Blair (1998), has stated that: "The days of the all-purpose (local) authority that planned and delivered everything are gone. They are finished. It is in partnership with others -public agencies, private companies, community groups and voluntary organisations - that local government's future lies. Local authorities will deliver some services but their distinctive leadership role will be to weave and knit together the contribution of the various local stakeholders ".
Urban freight movements and public-private partnerships
21
Erlach (2002) has distinguished the key differences between these two types of PPP (i.e. those with a narrow meaning and those with a broad meaning). The distinction between these two forms is that narrow PPPs are based on formal co-operation, whereas broad PPPs are based on informal co-operation. The differences are further illustrated in Table 1. Table 1 Differences between narrow and broad PPPs Narrow definition of PPP • Formalised co-operation venture
Broad definition of PPP joint • Relationship between partners is only partly formalised or not at all formalised • Resources made available by both • Partners retain control of the partners put at disposal of joint venture resources they provide • Risk and reward sharing • Information sharing • Co-ordination through joint venture • Co-ordination through network hierarchy structures • Applies mainly in the 'doing' phase • Applies mainly in the 'planning' phase Source: based on Erlach (2002). in
a
A narrow PPP is intended to involve the private sector in public projects. A broad PPP involves the public sector's intervention into private practices and operations, as well as consultation and dialogue in public decision-making. Though having different backgrounds, both types of PPPs can provide opportunities for the private and public sectors to benefit by sharing information and working towards common objectives. It should also be noted that a specific, narrowly defined PPP project could make use of a broadly defined PPP process during its planning stages to ensure a range of views and opinions are taken into account.
PUBLIC-PRIVATE PARTNERSHIPS IN URBAN DISTRIBUTION
The rationale for a PPP approach in urban freight transport Logistics activities are primarily performed by private companies. However, government (local and national) is expected to play a responsible role for many reasons - for example: • • •
coping with negative externalities such as road congestion and air pollution; necessary co-ordination with other public purposes such as city planning, regional economic development, environmental management, etc.; cross-border administrative issues with relation to international Supply Chain Management.
As discussed in the previous section, PPP can involve a range of interactions between public and private actors. In the field of urban freight transport PPP can take a variety of forms - for example: • •
private sector development of public infrastructure projects; operational agreements between parties (e.g. vehicle routings, delivery times, etc.);
22 • •
Logistics systems for sustainable
cities
consultation based on one party requesting the others views in written form (i.e. not a conversation but feedback to ideas, comments on a proposal etc.); two-way open conversation and dialogue about existing and future policies.
In some cases private companies are simply informed by policymakers about regulations. This type of interaction between the public and private sectors is not really based on partnership. National and local governments do not have a very good track record in involving urban distribution actors in decision-making in recent decades. Instead participation of such groups in policy-making has been often kept to a limited consultation exercise at best. However, this has begun to change in the last few years as interest in urban distribution has grown among policymakers and they have decided that a more inclusive approach is likely to result in more efficient and sustainable outcomes. For example, after perhaps twenty years of receiving little research or policy consideration in the UK, urban freight transport and distribution has recently begun to be recognised as an important activity by policy makers. During this period the UK central government published or said little about freight transport in general, and in particular about urban freight transport. However, this situation is now changing in the UK. Renewed interest in urban distribution issues among policy makers has been indicated by the establishment of a Freight Distribution and Logistics Unit in the Department for Transport, and the publication of the 1998 Transport White Paper "A New Deal for Transport: Better for Everyone" (DETR, 1998) and the daughter document to the White Paper entitled "Sustainable Distribution" (DETR, 1999). These documents outlined the UK government's determination to recognise and address the problems both faced and caused by distribution activity including those specifically concerned with urban freight movement. The urban freight transport and distribution considerations of local authorities in the UK have traditionally tended to take place as a reaction to problems, usually arising from complaints made by residents and other road users. Most local authorities with an urban remit have not developed coherent freight transport policies to the same extent that they have their public transport policies. However, local authorities are now being encouraged by central Government to focus greater attention on freight transport and to include consideration of urban distribution and its sustainability in their Local Transport Plans (LTPs). Similarly, during the last decade in the Netherlands, governments have become aware that cooperation with the private sector is very important in order to implement public policies. Government now seeks co-operation with the private sector and develops policies in full consultation with the private sector, in order to create win-win situations. This has meant that instead of regulation, local, regional and national governments now sign covenants with organisations representing business or directly with businesses. In these covenants the private sector agrees to behave in a particular way, while the public sector either provides facilities, finance, or reassesses and alters regulations. Platform Stedelijke Distributie (PSD or the Forum for Physical Distribution in Urban Areas) in the Netherlands is an example of this type of approach (see Box 3). Other examples include local government making arrangements with local retailers or transport companies. The policy agenda of PSD is developed in co-operation with both the public and private sector. The
Urban freight movements and public-private partnerships
23
implementation of the policy requires the public and private sector to work together in a partnership. The projects carried out by PSD are all examples of how this works.
24 Logistics systems for sustainable cities The Japanese national government authorised a set of policies for freight transport entitled "The New Comprehensive Program of Logistics Policies' in 2001, which was the revised version of the former program, first launched in 1997. Urban freight transport is considered an important area in which to achieve efficient and environmentally friendly logistics systems in Japan. Two quantitative targets were set on 'the load factor of trucks' and 'peak-hour average travel speed' in three major metropolitan areas; from the current 45 percent to a target of 50 percent, and from the current 21 km per hour to a target of 25 km per hour, respectively. In order to realise these targets, the program highlights the importance of co-ordination between public and private sectors, and between national and local governmental agencies, among others. This is why the program requested that the local agencies establish an independent organisation to plan local logistics policies, and new round tables to exchange information on local logistics policies inviting private representatives from the associations of carriers, retailers, etc. It can therefore be seen that participation plays an important role in these broad PPPs in urban distribution. The complexity of urban distribution can make it difficult to develop broad PPPs based on high levels of participation. Ogden (1992) argues that the urban freight system is far more complex and heterogeneous than urban passenger transport. This complexity and heterogeneity are driven by certain key features of urban goods movement, one of which is the range of participants involved in urban freight and the range of perceptions they hold of the "urban freight problem". Some are concerned with demand and most with some aspect of supply, they include numerous shippers, receivers, forwarders, freight and logistics companies, truck drivers, service companies, terminal operators, road and traffic authorities, government, and those living and working in urban areas who are affected by freight transport. Such complexity makes successful participation difficult to achieve.
Comparing the features of PPPs in urban freight movement Private companies (retail, wholesale or transport companies) carry out urban freight transport operations. The public sector is responsible for regulating and facilitating urban freight transport. Therefore a distinction has to be made between private and public strategies or measures. Table 2 shows a classification of strategies or measures. Public measures are actions taken by public authorities and are intended to bring about behavioural changes in the private sector. The public sector can involve the private sector in the creation and development of these measures through consultation and dialogue. Private strategies, such as voluntary cooperation between companies, are initiated by the private sector without public sector involvement. Some strategies and measures involve the direct participation of both the public and private sectors. Technology improvement in fields such as road and traffic information, and the development of new vehicle standards are examples of public-private measures and strategies.
Table 2 Classification of public and private measures (examples) Policy measures and instruments Applied on
Public
Private
Licensing and regulations Zoning for logistics activities or transport intensive retail Minimal load-factor
Pricing
Financial support
Land use pricing
Subsidies for land use prices
Networks
Truck routes, vehicle and time restrictions
Road pricing
Terminals Loading/unloading
Urban distribution centre Loading time
Vehicles
Emission standards
4/ Land use
Logistics operation
Subsidising intermodal transport New infrastructures for freight Terminal exploitation Differentiated parking charges Fuel taxes
Source: Visser, Binsbergen and Nemoto (1999)
Voluntary cooperation Concentrate businesses on one location Load exchange
Facility support Subsidies for low emission trucks
Operation of terminals Sharing unloading facilities Share of vehicle fleet
Public and private Technology improvement
Information systems
New load-units
Cargo information systems
Road construction
Real time traffic information
Transhipment and storage Off-street unloading facilities Electric vehicles, handling equipment
Reservation system of parking lots Vehicle tracking systems
26
Logistics systems for sustainable cities
An important aspect of measures concerns the problem of adoption. The adoption of measures can be supported by making the desired behaviour more attractive (financial support and licensing) or by discouraging other behaviour (pricing and regulation).
EXAMPLES OF URBAN FREIGHT INITIATIVES THAT MAY INVOLVE PPPS The following section examines several of the PPPs in urban freight that were highlighted in Table 2 in greater detail. Firstly urban transhipment centres and intermodal centres are discussed, followed by alternative power vehicles, Freight Quality Partnerships, intelligent transport systems, Low Emission Zones and congestion charging. Urban transhipment and consolidation centres Transhipment centres are frequently suggested as a solution to the environmental problems caused by lorry traffic in urban areas. In such an approach, freight destined for urban areas is unloaded at a depot on the periphery and transhipped into small vans for final consolidated delivery. These vans also collect consignments from city centre premises. Proposals may envisage compulsory use of such facilities, with all other lorries banned from a designated area, or they may be more voluntary in nature, hi the latter case various incentives may be employed to promote their use. In addition, operators choosing not to use the facilities may face severe time-of-day or vehicle size restrictions imposed by local authorities within the urban area (Ogden, 1992). Proposals were developed in the 1990s to establish transhipment systems in a number of Dutch towns and cities following consultancy studies of their potential use and cost effectiveness. Experimental schemes were proposed in four cities. The first such experiment eventually got under way in Maastricht in the early 1990s but the volumes going through the depot were low. Progress on schemes for other cities was hampered by problems in agreeing the precise nature of these schemes. Who should own the facilities - the public sector or private enterprise? Should their use be voluntary or compulsory? What sort of licensing system should be put in place for operators involved in the collection and delivery work in the area concerned? What restrictions should be placed on vehicle size, type and hours of operation for operators remaining outside the scheme? Despite these problems there is still interest in transhipment centres as a potential solution. Several UK local authorities have investigated their feasibility in recent years, though none has progressed beyond the initial investigation stage. In France, a scheme got under way in 2001 in the historic town of La Rochelle, where narrow cobbled streets in the town centre are not suitable for large vehicles (BESTUFS, 2002). The most significant problems facing such schemes appear to be the relatively high costs of the transhipment operation and the loss of control suffered by the shippers of the goods. Historic city and town centres are clear examples where there may be an argument to develop a consolidation centre. However, there are other experiments that are also relevant. For example, the development of a consolidation centre at Heathrow airport was stimulated by the desire of the airport operator (BAA) to reduce the number of goods vehicles entering the airport to deliver to the extensive terminal retail businesses. The consolidation centre has resulted in
Urban freight movements and public-private partnerships
27
significant reductions in goods vehicle movements within the terminal area and also had beneficial environmental impacts (Department for Transport, 2002). In Japan four public distribution centres were built in the outskirts of downtown Tokyo in the 1960s by a national government-affiliated corporation. Each centre, approximately 100 hectares in size, consists of truck terminals, container depots, warehouses, wholesale markets, and shops for carriers and wholesalers. They had a number of rules to control the usage of the centres' space, which discouraged the carriers from locating their facilities there. As a result, the carriers built their own centres independently on sites that were not always suitable for that purpose because of environmental impacts on residential areas. In 1985 this corporation was fully revitalised in order to meet rapidly changing carriers' requirements efficiently. This was achieved through a legal change of status although more than 25 percent of the stock of the corporation is still held by the local government. This reflects the growing recognition that logistics policies are required to meet both public and private interests in a flexible and intelligent way.
Intermodal centres In addition to the transhipment centres discussed above, some freight centres have wider objectives than simply transferring goods from one type of road-based vehicle to another for final delivery in the urban area. Freight centres can also be intended to boost the regional economy, and enhance international trade. Such freight centres will be equipped with modal interchange facilities (e.g. road-rail, road-rail-sea, etc.) and often also include stockholding facilities. Various terms exist to describe these types of centres including 'freight villages', 'special logistics areas', and 'logistics parks'. Some of these intermodal centres are located in large-scale industrial and business parks, providing services to the companies based there. For example, Gilterverkehrszentrums (GVZ's or Cargo Traffic Centres) have been developed in Germany since the early 1990s. These centres are intended to create inter-regional networks between conurbations. This is an initiative taken by the national government, and the objective is to develop 30 GVZ locations that are capable of shifting traffic from roads to rail and ship. GVZs have already been developed in Bremen, Augsburg, Dorpen, Dortmund, Hannover, Leipzig, Milrichen, Neurenberg, Rostock and Trier. Given the size and infrastructure requirements of intermodal centres, at the very least their development involves close contact and planning between the public and private sectors, hi some cases these centres are either owned by, or the development is financially supported by national, regional and local authorities.
Alternative power vehicles One potentially significant advantage of transhipment centres (discussed above) is that they can be used in conjunction with other measures to generate wider benefits. Sites adjacent to railway lines and waterways may be chosen to maximise the scope for inter-modal operations, for example. Transhipment strategies can also be linked to relatively severe time-of-day or lorry weight restrictions in city centres, as explained above. Perhaps their most important advantage is that because the fleet of vehicles based at the centre is dedicated to urban collection and delivery work, such vehicles can be specified most appropriately for the town or city concerned.
28 Logistics systems for sustainable cities Attention can be paid to the most suitable vehicle size, and more environmentally friendly vehicles, perhaps with quieter engines or powered by gas or electricity, can be used. Assessment of electric powered small delivery vehicles is in fact one of the objectives of the La Rochelle scheme outlined above. There seems little doubt that the use of environmentally friendly vehicles will increase, particularly if tax inducements for alternative fuels and for cleaner and quieter engines are stepped up and if alternative fuels are made more readily available. At present, technologies for alternative fuels and quieter operation are relatively new and vehicles incorporating such technologies are comparatively rare. As a result, they are more expensive to buy. hi the UK, the highly competitive nature of the transport and logistics industry may be holding back the introduction of such vehicles, given their high prices at present. Operators need to be reassured that lower fuel prices due to tax concessions will be maintained into the future, to allow payback on their capital outlay. Operators might also be more easily persuaded to change fuels if there was more guidance available on which of the various alternative technologies (electric, gas, fuel cell, biomass etc) are likely to become generally adopted in the future. A UK example well publicised a few years ago is the use of natural gas powered vehicles by BOC Distribution Services (now part of Gist) on their dedicated contract to supply Marks and Spencer outlets in central London (Distribution, 1997). There appears to have been more interest in environmentally friendly urban freight vehicles on mainland Europe than in the UK. Some Scandinavian and German cities have experimented with low noise and low emissions vehicles, for example in Heidelberg. ELCIDIS, an EC THERMIE project, has established demonstration sites in three large European cities (Rotterdam, Stockholm and Milan) as well as three smaller cities (Erlangen and Stavanger in addition to the case of La Rochelle mentioned earlier) for the testing of electric powered distribution vehicles. Interest in the UK is likely to increase, however. Air quality is routinely monitored in UK cities, and in London, where air quality is the worst in the UK, the feasibility of a Low Emission Zone is under investigation. Whilst such a zone will not be implemented until 2005 at the earliest, such proposals will encourage operators to seek out and evaluate low emissions technologies. Following on from the arguments set out in previous sections, it is more likely that operators will specify environmentally friendly vehicles if such vehicles can be dedicated to urban work. Freight quality partnerships Freight Quality Partnerships (FQPs) are an approach launched by the Freight Transport Association (FTA) in 1996. The FTA initiative brought together industry, local government and representatives of local and environmental interest groups to pursue the following agenda (FTA, 1998): • •
To identify problems perceived by each interest group relating to the movement and delivery of goods in their city; To identify measures within the group's competence to resolve or alleviate such problems;
Urban freight movements and public-private partnerships •
29
To identify best practice measures and principles for action by local government and industry to promote environmentally sensitive, economic and efficient delivery of goods in towns and cities.
The FQP initiative was tested in four UK urban areas in 1996: Aberdeen, Birmingham, Chester and Southampton. The UK Government has been promoting FQPs since 1999 (DETR, 1999). FQPs can facilitate improved dialogue about urban freight transport issues between local authorities, freight transport companies, retailers, manufacturers and other businesses, local residents and other interested parties. This can lead on to more efficient, less harmful operations. In their guidance document the government states that, "Freight Quality Partnerships provide local authorities with a means to formalise the consultation and development work undertaken in their sustainable distribution strategy. Authorities have an integral role to play in helping industry, through developing partnerships to progress and develop best practice in sustainable distribution systems, and to find solutions to the issues of greatest concern. For example, freight quality partnerships provide a good means of delivering air quality and noise benefits while removing peak hour traffic and improving the efficiency of deliveries at the same time. Companies can be given improved access to premises and extended delivery hours, including night time deliveries, in return for agreeing to use cleaner, quieter vehicles and agreeing a night time code of practice" (DETR, 2000). Approximately 50 local authorities referred to the development of FQPs or similar schemes under a different name in their Local Transport Plans (LTPs). However, study of the LTPs that mention FQPs shows that there are significant differences in how these local authorities are choosing to define FQPs, and some are still in the process of working towards the introduction of FQPs rather than setting them up now. The LTPs indicate that approximately 30 local authorities have already put in place formal agreements and arrangements for a FQP. These authorities include: Hampshire, Southampton City, Surrey (FQP established in Guildford), Kent (FQP in Canterbury), Ripon, Northamptonshire, the West Midlands, Leicestershire, and Nottinghamshire. FQPs have been established for a number of purposes ranging from Regional Strategic Partnerships, to city- or town-specific partnerships, to micro-level partnerships (maybe concerned with a few streets), to issue specific partnerships. UK Government guidance suggests that FQPs should try to involve representation from logistics companies, retailers, manufacturers, service providers, rail operators and the local airport or sea port. In addition, the government suggests that other potential representatives including the Chamber of Commerce, the police, environmental groups and resident groups should also be involved (Department for Transport, 2002). FQPs should help ensure freight and service transport receives the level of attention it deserves, providing recognition of the fundamental role played by freight and service vehicles in the functioning of towns and cities. FQPs should play an important role in finding a suitable balance between economic and environmental pressures in UK urban areas. However, there are several unresolved issues concerning FQPs. These include: • •
How to include freight and service companies not based in urban areas; How to involve a significant proportion of all relevant companies;
30 Logistics systems for sustainable cities • •
The level of public funding available for policy measures, initiatives and enforcement; How to ensure compatibility between policymaking at the local, regional and national levels.
Use of intelligent transport systems There is significant scope to improve the efficiency of logistics operations through the greater use of information technology. Transport modelling work reported in Taniguchi, Thompson, Yamada and Van Duin (2001) has demonstrated that effective use of dynamic vehicle routing and scheduling systems can produce significant benefits in terms of both economy and the environment. In-cab information systems and mobile data systems allow operators to save time and money by advising drivers on how to avoid congestion. Electronic proof of delivery systems, as used increasingly by express parcels companies, can reduce the time parked outside customers' premises. Information technology may also facilitate voluntary consolidation schemes. Operators willing to co-operate on the German 'city logistics' model could use real-time information systems to track consignments destined for the city centre, identify operators with spare capacity to handle such consignments and route them accordingly. Electronic tagging and scanning of consignments facilitates traceability throughout the supply chain, which may allay shippers' fears over the loss of control at transhipment centres. An interesting experiment has been conducted in Osaka in western Japan using electric vans and intelligent transport systems (Taniguchi and Nemoto, 2003). The experiment involved 79 companies that participated voluntarily. These companies can use 28 small electric vans that they book through the Internet in advance. The vans are kept in eight parking places ready to be picked up. The users can return the vans to any of these parking places after using them, so that the users can avoid driving the empty vans on the congested roads and instead can take a subway when they come back to their offices. Each van is equipped with a mobile data-communication system and GPS (Global Positioning System) to identify the vehicle location. These systems allow the Centre to control the vans including; booking and renting vehicles, operation and route guidance, and management of the balance of electricity. The experiment has been conducted successfully without any serious technical problems. A questionnaire survey for frequent users shows that the benefits of reduced travel time were limited to 33 percent of them, partly because they often return the vans to the same parking places, contrary to the study team's expectation. It is interesting that 31 percent said that they enjoyed the reduction of travel time using car navigation systems. About half of the users were willing to pay 300-600 yen per hour, which was too low as compared with costs for operating the system. Subsidies are required to make the system economically feasible at present.
Low emission zones The aim of a Low Emission Zone (LEZ)/ Environmental Zone is to improve air quality by excluding older, high-polluting vehicles from specific urban areas and encouraging the faster take up of more modern, cleaner vehicles.
Urban freight movements and public-private partnerships
31
Environmental zones were implemented in the central areas of Stockholm, Goteborg, Malmo, and Lund in 1996. Diesel driven trucks and buses with a gross vehicle weight of over 3.5 tonnes are only allowed to enter these environmental zones if their engine is less than 8 years old or achieves specified emissions standards (Municipality of Stockholm et al, 2002). A similar scheme exists in central Amsterdam in which vehicles over 7.5 tonnes gross weight are only allowed to enter if they meet specified emissions and size criteria and also have a load consolidation of at least 80% (PSD, 2002). A feasibility study for the introduction of a LEZ in London has recently been carried out (ALG/GLA, 2002). This was jointly commissioned by the Mayor of London, the Association of London Government and two central government departments to investigate ways of improving the quality of air in London. This is in response to recent UK and European legislation that has introduced target levels for air quality in forthcoming years. If the London authorities decided to go ahead with an LEZ for London the earliest possible date it is likely to be implemented is 2005. An important part of the London feasibility study was to consider how the introduction of an LEZ in London would affect the businesses, organisations and communities that live and work in the London area and to ensure that any scheme would be publicly acceptable. The study team therefore contacted more than 500 stakeholders and interviewed approximately 100 stakeholders including business representatives, local and central government, freight industry associations, bus and coach organisations, taxis representatives, and environmental and transport NGOs to get their opinions on the LEZ concept and their views on implementing the enforcement strategies. Interviews were also held with goods vehicle operators in order to obtain their views on the LEZ concept, their likely behavioural responses, and the likely impact on their distribution operations and costs when working in London. The Mayor of London, the Association of London Government and the two central government departments that commissioned the research are currently considering the findings of the feasibility study and deciding the appropriate action to take with respect to a London LEZ, taking into account the views expressed by stakeholders.
Congestion charging Congestion charging refers to a scheme in which vehicle drivers (or the companies responsible for the vehicles) have to pay a charge in order to enter a particular geographical area at a particular time. The aim of such a scheme is to reduce road traffic levels in the urban area and also to reduce traffic pollutant emissions. Such a scheme may also generate a profit which can be used to provide improved public transport services. Congestion charging was implemented in London in February 2003. In the scheme drivers will pay £5 per day to enter central London between 07:00 and 18:30 from Monday to Friday. Goods and service vehicles working in central London are all subject to this charge. The Mayor of London anticipates that the congestion charge will reduce traffic levels in London, and that freight and service companies will benefit in terms of shorter and more reliable journey times.
32 Logistics systems for sustainable cities It was originally proposed in the London scheme that goods vehicles should pay £15 per day. The freight industry was critical of this charge. The proposed charge was subsequently reduced to £5 per day for all vehicles. Prior to the introduction of the scheme, Transport for London anticipated that congestion charging would result in 10-15% reductions in traffic levels, with speed improvements of 1015% inside the zone (TfL, 2002). Clearly, reductions in traffic could lead to greater reliability for the journey times of goods vehicles. Increased reliability would off-set some or all of the additional costs but there remains some uncertainty about the likely impacts. In addition, while it is argued that traffic would fall in the congestion charging area it has also been claimed that congestion would be worse around the edge of the zone. This in turn would reduce the level of benefits to be expected from more reliable delivery and service trips in the central area. Consultation with stakeholders and the general public took place at several stages during the development of the congestion charging scheme. This began in 2000, when the Mayor, Ken Livingstone, set out his initial ideas in a document and sent this to key stakeholders including local councils, businesses and road user representatives to better understand their views on congestion charging and transport issues in London. The comments provided by these stakeholders helped shape the Mayor's draft Transport Strategy, published on 11 January 2001. Public consultation on the congestion charging scheme took place between July and September 2001. This involved the following activities (TfL, 2002): (a) (b) (c) (d)
A public exhibition A 12 page public information leaflet was produced Two large public meetings were held A call centre was in operation throughout the duration of the consultation (it received over 2,500 calls from individual members of the public) (e) Information about the congestion charging scheme was placed on the TfL Street Management website (f) Advertisements were placed in newspapers and on the radio to inform Londoners of the consultation. Analysis of responses were carried out and some scheme modifications were proposed as a result of this consultation process. TfL produced an information pack outlining these proposed modifications. This pack was made available for public inspection at the offices of TfL Street Management and at eight London Boroughs. In addition, the pack was also sent to 500 key stakeholders and TfL arranged a series of consultation meetings with key stakeholders. The Mayor had the power to decide whether or not to hold a public inquiry into the congestion charging scheme. He chose not to hold an inquiry. The consultation process that took place as part of the congestion charging scheme was a traditional approach to consultation, providing companies, organisations and individuals with an opportunity to submit their opinions but with no commitment from the Mayor to include them in the decision-making process. In June 2001, Tokyo Metropolitan Government (TMG) released a report to advocate the introduction of road pricing in central Tokyo (Tokyo Metropolitan Government, 2001). Four alternative charging areas were proposed based on combining major ring roads, rivers and
Urban freight movements and public-private partnerships
33
railways as the cordon line. A car crossing the line into the area has to pay the charge. A promising alternative area covers 72 km2, which is larger than central London (21 km 2 ). The charging period proposed was from 07:00 to 19:00 on weekdays. hi order to clarify the effects of road pricing, TMG conducted two surveys to interview passenger-car drivers and managers of trucking firms. Assuming 500 yen and 1,000 yen charges for small and large vehicles respectively, 17 percent of small commercial trucks and 10 percent of large commercial trucks showed that they would reduce truck use by improving the utilisation ratio of their fleet and the loading ratio of each truck. This would be achieved with more efficient fleet management and a co-operative delivery system. The results indicated that road pricing would affect private (own-account) trucks with relatively low loading factors most significantly, with 30 percent of these respondents answering that they would change their truck use. It implies that cargo would shift from the private trucks to the commercial (thirdparty) ones to some extent. Though road pricing has social objectives to reduce road congestion and to improve environmental situation, it would also affect the stakeholders differently. In order to introduce the scheme successfully with public acceptance, TMG has started consultation with stakeholders and the general public since 2002. TMG's consultation efforts, however, are not as intensive as those in London. For the Mayor of London, road pricing was an important campaign issue and this has been reflected in the rapid development and initiation of the scheme. The TMG Governor Ishihara, by contrast, appears to have stepped back without any clear message on road pricing in the recent re-election campaign in May 2003. It would appear that, from a political perspective, road pricing seemed unlikely to increase his support. It is clear that in order to make major changes to urban transport, strong leadership is an important factor in the successful implementation of schemes involving broad PPPs.
CONCLUSIONS It is too early to evaluate the success of narrow financially-driven PPPs (i.e. particular projects in which the public and private sector have shared interests and objectives and where there is often an element of shared risk and reward). The benefits of this approach are that it reduces public sector capital investment and makes use of private sector expertise and project management skills. It has also been argued that this approach provides financial incentives to finish work on time and in budget and to achieve stated targets. However, there is currently uncertainty about whether these narrow PPPs offer good value for money (i.e. whether they cost taxpayers more in the long-term than straightforward public sector investment). It is argued that the government could borrow the capital more cheaply on the private capital markets itself. In terms of broad PPPs (i.e. urban distribution initiatives between the public and private sector that involves information dissemination, communication, co-operation or joint working) this approach should help to ensure that freight transport receives the level of attention that it deserves in urban areas. While traffic levels and their impacts in towns and cities have received growing attention in recent years, much of this has been directed at public transport and private car traffic with relatively little consideration paid to road freight transport. The goal is to find a suitable balance between economic and environmental pressures.
34 Logistics systems for sustainable cities There are several issues concerned with broad PPPs that still have to be resolved. These include how best to include freight transport companies that are not based in the urban area but operate vehicles in the area in discussion and consultation processes (for example, this has proved difficult to achieve in the FQPs being developed in the UK). Another concern is ensuring that different towns and cities do not implement local measures that, although efficient at a local level, are inefficient at a regional or national level - for example different types of vehicle requirements and restrictions in different towns and cities may increase total fleet requirements and trip numbers. Hopefully involvement of larger freight transport companies that operate in many different urban areas should help to prevent this, together with co-ordination between different levels of the public sector. As the examples reviewed in this paper indicate, it is important to bear in mind that effective broad PPPs take some time to establish and they also take time before they begin to yield results. "Effective solutions to most freight transport problems require substantial cooperation between the private sector, where goods are moved, and the public sector, which provides and maintains the roadway system infrastructure. That's easy enough to say, but in reality such co-operation requires a degree of credibility and trust which takes time and effort to build" (Millendorf, 1989). Bearing this in mind, policy makers need to be clear about the issues they want to engage the private sector in consultation and joint working on, and to decide how best to use the time and efforts of the private sector in these initiatives. Focusing on the key issues and outcomes will help to engage and retain the private sector's involvement in such initiatives. Given the wide range of stakeholders involved in freight transport considerations in urban areas (including retailers, wholesaler, carriers, warehousing, residents, shoppers and workers) it will undoubtedly prove difficult to both engage and please everyone. However, if the focus remains on ensuring that the delivery and collection of goods in urban areas takes place as in an efficient manner, while imposing as few social and environmental impacts as possible there are clearly benefits to be achieved through the use of a broad PPP approach.
REFERENCES Association of Local Government and the Greater London Authority (ALG/GLA) (2002). London Low Emission Zone Feasibility Study: Phase I Report of the Steering Group, http://www.london-lez.org/ BESTUFS (2002). Land use planning and business models for urban distribution centres, BESTUFS workshop, 25-26 April 2002, La Rochelle, France, http://www.bestufs.net/ Blair, T. (1998). Leading the Way: A New Vision for Local Government, IPPR, London. Bockel, R.van. (2002). PSD - Public Private Partnership(s) in The Netherlands For Urban Freight Transport, presentation at BESTUFS workshop on Public Private Partnerships, 12-13 September 2002, Malaga, Spain, http://www.bestufs.net Currie, D. (2001) quoted in Scott. J, (2001). h PFI a good deal?, 3 September 2001, article on the BBC News website, http://news.bbc.co.uk/l/hi/in_depth/business/2001/ppp/1496562.stm Department of the Environment, Transport and the Regions (DETR) (1998). A New Deal for Transport: Better for Everyone, The Government's White Paper on the Future of Transport, Cmnd.3950, The Stationery Office.
Urban freight movements and public-private partnerships
35
Department of the Environment, Transport and the Regions (DETR) (1999). Sustainable Distribution: A Strategy, DETR. Department for Transport (DfT) (2003). A guide on how to set up and run Freight Quality Partnerships, Transport Energy Good Practice Guide 335, DfT. http://www.freight.dft.gov.uk/gpg355/pdf/gpg335-final.pdf Department for Transport (2002). Heathrow Airport Retail Consolidation Centre, Transport Energy Good Practice Case Study 402, DfT. http://www.transportenergy.org.uk/vpo/downloads/letter/GPCS402.pdf Distribution, (1997). M&S goes for Gas, Distribution, Vol.10, No.5, October, p.2. Erlach, P. (2002). Public Private Partnerships in Austria, paper made available as part of BESTUFS workshop on Public Private Partnerships, 12-13 September 2002, Malaga, Spain, http://www.bestufs.net Freight Transport Association (1998). Delivering the Goods, FTA, Tunbridge Wells1 HM Treasury (2000). Public Private Partnerships - the Government's approach, The Stationery Office, London. Kelly, G. (2001) quoted in Scott. J, (2001). Is PFI a good deal?, 3 September 2001, article on the BBC News website, http://news.bbc.co.uk/l/hi/in_depth/business/2001/ppp/1496562.stm Lowndes, V. (2001). Local partnerships and public participation, paper prepared for publication by IPPR Partnerships Commission, London. Millendorf, S. (1989). Facilitation of goods movement in the New York City area, in Chatterjee, A., Fisher, G., and Staley.R. (eds) Goods Transportation in Urban Areas, American Society of Civil Engineers, pp.25-30. Municipalities of Stockholm, Goteborg, Malmo, and Lund (2002). Environmental Zone for Heavy Traffic: Regulations in Stockholm, Goteborg, Malmo, and Lund. Ogden, K. (1992). Urban Goods Movement: A Guide to Policy and Planning, Ashgate, Aldershot. Pretty, J. (1995). Participatory learning for sustainable development, World Development, Vol.23, No.8, pp.1247-1263. PSD (2002). Personal communication from PSD (the Forum for Physical Distribution in Urban Areas) The Hague, the Netherlands. Sardi, L. (2002). Public private partnership in the city of Parma, presentation at BESTUFS workshop on Public Private Partnerships, September 2002, www.bestufs.net Taniguchi, E., Thompson, R.G., Yamada, T. and Van Duin, R. (2001). City Logistics: Network Modelling and Intelligent Transport Systems, Pergamon/Elsevier, Oxford. Taniguchi E. and Nemoto T. (2003). Transport-demand management for freight transport, in Taniguchi E. and Thompson R.G. (eds) Innovations in Freight Transport, WIT Press, UK. Tokyo Metropolitan Government (2001). The Report of TMG Road Pricing Committee, (in Japanese). Transport for London (TfL) (2002). The Greater London (Central Zone) Congestion Charging Order 2001: Report to the Mayor, February 2002, TfL. Visser, J., Binsbergen, A.van. and Nemoto, T. (1999). Urban freight transport policy and planning, in Taniguchi.E, and Thompson.R, (eds), City Logistics I, Institute for City Logistics, Japan, pp.39-70. Whiteing, A. E., Browne, M. and Allen, J. (2003). City logistics: the continuing search for sustainable solutions, in Waters, D. (ed.) Global Logistics and Distribution Planning: Strategies for Management, Kogan Page, London, pp.308-320.
This page is intentionally left blank
3
TRANSPORT DEMAND, TRANSPORT AND TRAFFIC FLOW - KEY ELEMENTS OF CITY LOGISTICS Raluca Raicu, Transport Systems Centre, University of South Australia, Australia Serban Raicu, Polytechnic University of Bucharest, Rumania
ABSTRACT The authors were motivated in addressing the subject of the current paper by the too frequent misuse of the following notions - transport demand, transport flow, and traffic flow - and by the vast amount of literature dedicated to describe the four step demand models. Is it justified to attribute reference value to these models for dimensioning the transport supply in the urban freight distribution? Could demand models contribute to the estimation of the operation parameters associated to the transport means in the conditions of a public infrastructure on which other transport flows exist (apart from those that made the object of the modeling)? The paper tries to answer these questions and to clarify some other related issues. In this context, the authors attempt to make a complex and interconnected description of the transport demand, transport and traffic flow. Criticising the simplest and most traditional measure associated with demand - the one resulted by assimilating demand with service volumes - a complex description of the demand is proposed. Thus, transport demand is physically characterized by transport links, value, and temporal characteristics. On the other hand, transport demand is commercially characterized by reliance on the quality of service, the level and structure of the transport system load, reliance on the mode split and route choice in the conditions of a unique, multimode transport market, with interchangeable, complementary or independent services. Same as in the case of transport demand, the definition of transport flow characteristics goes beyond the traditional framework. The authors propose physical characteristics (value,
38
Logistics systems for sustainable cities
structure, temporal characteristics) and functional characteristics (homogenous/ heterogeneous flow, directional equilibrium/variability, single mode, combined or multimodal assignment, quality parameters according to the logistics requirements). The superposition of transport flows over the infrastructure produces the traffic flows. A complex characterisation of these is proposed too. From this research results that the ex-ante, ex-post transport demand, the transport and traffic flows are different notions. For their description one needs a multitude of physical, commercial, technological, environmental parameters to allow precise and complete identifications. Transport demand (ex-post) influences the characteristics of the transport flow which, in turn determines the characteristics of the traffic flow, but the dependencies are not direct. The specific influences of the demand, transport means, infrastructures and technologies make difficult the representation of these dependencies with the help of analytical models. Therefore, in order to evaluate and predict the transport flow dynamics and specially the traffic flow dynamics is advisable to use experiments and empirical estimations. The quantitative measures proposed to describe transport demand, transport and traffic flow and the diagrams which represent the links between them as well as the links with the elements of the transport system in the opinion of the authors bring the necessary details required for modelling the urban freight logistics. The outcomes of this research are useful to urban logistics both in the development of evaluation techniques and policy making.
INTRODUCTION Tanaguchi et al (1999, 2001) defined City Logistics as, "the process for totally optimising the logistics and transport activities by private companies in urban areas while considering the traffic environment, the traffic congestion and energy consumption within the framework of a market economy". The objectives of this city logistics optimisation are of interest for the both public and private sector. The four actors involved in urban freight transport: shippers, freight carriers, residents (consumers), and administrators have their own objectives and act in different manners. Consumers and shippers attitudes are mainly dictated by the characteristics of the transport demand; that of freight carriers and shippers by the transport flow characteristics, and that of administrators and consumers by the traffic flow parameters. At the same time, the decisions of the four actors influence the characteristics of transport demand, and specially those of transport and traffic flows. These decisions determine the quantitative and qualitative level of accomplishment for freight transfer in urban environment, the quantity of resources required (transport means, loading-unloading equipment, warehouses, transport infrastructure, fuel and energy, human resources, etc), the social and environmental effects. Transport demand, transport and traffic flow reflect the way in which the transport system has adapted its infrastructures, transport means and technologies to the requirements of city logistics. The response of the transport system to the requirements addressed by the socioeconomical environment, in the form of transport demands, consists of convergent solutions for the satisfaction of logistics constraints with a minimum (rational) use of resources. From
Transport demand, transport & traffic flow - key elements of city logistics 39 these demands derive the transport flows and then the traffic flows. This is why we believe transport demand, transport and traffic flow represent key elements for the city logistics. But what are the correlations between these elements? Are they used adequately? Are they correctly characterised and are the specific dissimilarities exposed? Is it justified to attribute reference value to the four step demand models (Ortuzar and Willumsen, 1994; Oppenheim, 1995; Slinn et al, 1998) for dimensioning the transport supply (infrastructure, transport means) in the urban freight distribution? The paper tries to answer these questions and to clarify some other related issues.
TRANSPORT DEMAND It is common knowledge (Raicu and Popa, 1997), that irrespective of the type of demand (technological-internal, commercial-external) one has to distinguish between the ex-ante demand, D a , (the notion of need, representation of the aspirations of the beneficiaries before the constraint appears) and the ex-post demand, D p (demand satisfied by the concrete transport system). Both a quantitative and qualitative estimation of the ex-ante demand is necessary using physical, mathematical and statistical models. The quantitative estimation of transport flows and implicitly of transport demand always introduces unknowns related to the operational conditions of the network. For instance, in the case of passenger transport, the travel time is better than distance for determining the flow between two sub-systems (the case of gravity type models for example). But, travel time is dependent on the network and the operational technology adopted (public transport routes, frequency, timetables integration, etc.). The estimation of the ex-ante demand has to depend as little as possible on the characteristics of the supply; it has to relate more to the system than to the network or the operational technology adopted. The demand-supply correlation, through its dynamic character, shifts the ex-ante transport demand towards the ex-post, stimulating or inhibiting the diversification and amplification of the ex-ante demand (Figure 1).
Figure 1 The link between ex-ante and ex-post demand Limiting the analysis to freight and personal commercial transport it is to be noticed that the simplest and most traditional measure associated to demand is the one resulting from assimilation with the volume of service. This is expressed in net tonne kilometres and passenger kilometres, with the advantage of being easily summed and partially linked to the transport costs and the available supply, but with the disadvantage of having no direct link with the demand and not defining it completely.
40 Logistics systems for sustainable cities In characterising the transport demand there are of course aspects specific to freight and passenger transport. The following analysis concentrates mainly on freight transport. a) Physical characteristics, D F : (i) D FR , relational characteristics: if I = 1, 2,....I, (I > 2) are origins and destinations, then between the elements of set I = {1,2,...I} we'll have transport links (routes, itineraries) which can be described by the pair (i, j), where the first element is the origin and the second element the destination. The set of pairs (i, j) can be described as R, R = {(i r , j r )} r _] • and together with the set of origins and destinations, I, define the demand graph (I, R) with I vertices and R arcs. It is on this incomplete, oriented graph with I > 2, R c l x l and — < R < 1(1 -1) that the quantities to be transported have to be specified, (ii)
(iii)
DFC, quantitative characteristics: the freight quantities to be transported, Q[- in a certain link (i, j), from the goods w = 1,2,K , W , with specific transport requirements according to their state (solid, liquid, gas) and their fragility, dangerousness, dimensions, etc. DFT, temporal characteristics: can be classified as transport demands with or without due date, periodical, or non-periodical for which the following have to be defined: 9j-
- the initiation time of the first transport from the examined series, T^.w' - the
transport frequency (periodicity), t>. tj.
- the imposed time for transport initiation,
, tj. - the earliest, respectively the latest time for delivery on (i, j) link for the type
w demand and the quantity Q [ W ) , or the average intensity X(;w) = Q j W ) / [ t [ w ) - t[. w) ]. (Raicu, 2002) b) Commercial (market) characteristics, Dc, to be attached to the physical characteristics above; they refer to the interaction between the assembly of transport demands and the supply of a transport mode or the assembly of transport supply: (i)
reliance on the quality of service — travel time, safety, convenience, transfer and/or delivery conditions - which, expressed in monetary value, added to the current cost of transport (fees from the users) determines the generalised cost, Cg(S), specific for a certain supply S
(ii)
reliance on the level and structure of the transport system load, because the generalised cost is not determined only by the supply, but also by the level and structure of the load, N, of the transport system (mode) elements
(iii)
reliance on the mode split and assignment (route choice); the transport market is a unique and multimode assembly of interchangeable (competitive), complementary or independent services; in this market, the global transport demand is distributed according to the supply the most suited to the demand characteristics, and according to
DCJ=f(Cg(S))
£>„ =
Transport demand, transport & traffic flow - key elements of city logistics 41 the freight carriers and/or beneficiaries preoccupations for avoiding congestion which otherwise substantially degrades the qualitative indicators of the operation Dc
where e m
m
3
= y/(em
m
)
is the cross elasticity of demand (the demand associated to the transport
mode mi, relative to the transport mode mj); based on this, competitive, complementary or independent relations are established between the transport modes. The increasing requirements of city logistics (e.g. correct quantity, just in time delivery, right place, preset price, unchanged characteristics of the goods carried) induce essential modifications in both physical and commercial demand characteristics as outlined above. Some of these are: 1. The freight categories w = 1, 2, ..., W, are established by promoting the logistic families of products in which the grouping is performed as to the transfer characteristics (distribution itineraries, ways of commercialisation, handling restrictions, storage, transport, size and frequency of the orders, admissible lateness in delivery, etc) and not as to the traditional, commercial ones (food, clothing, textiles, cosmetics, chemicals, etc). 2. The temporal characteristics are changing by properly dimensioning the stocks, by the emergence of intermediate warehouses, distribution centres, loading/unloading centres. 3. The transport demands become multimode and dependent on the generalised cost, by increasing the quality of service and reducing the transport cost; they are capable of influencing the a priori estimates (before the supply evaluation) concerning the demand market characteristics. 4. The transport demands, assigned to transport modes and itineraries, due to the grouping objectives of city logistics, change the area of technical and efficiency capabilities of the freight carriers. 5. The emergence and development of urban underground freight transport essentially modifies the demand characteristics discussed before. The problem of grouping the deliveries is no longer posed (as oppose to the present situation), but space-use problems arise - spatial activity concentrations and distribution centres, and at the same time a rise in public transport accessibility. Modifications of the transport demands (as outlined above) in the city freight logistics context are due to the correlated approach of the involved actors. They focus on global optimisation by taking into account the costs and benefits of the logistics schemes for both public and private sectors. Therefore, the transport demand characterisation has to be complex. This is accomplished by successive iterations considering the dynamic equilibriums between demand and supply specific to the urban environment.
TRANSPORT FLOW Transport demand applied to a certain transport system in relation to the available transport means, infrastructures, and technologies of the system becomes transport flow. In Figure 2, the transformation of the transport demand into transport flow, FT m , corresponding to a transport mode mj and to a network N, defined by the (K, D) graph, can be followed, where: M - transport means that operate for mj transport mode; T -technology of transport mode m;; N - infrastructure of transport mode mi; D F R; D F C ; DJT - relational, quantitative and
42 Logistics systems for sustainable cities temporal characteristics of the ex-ante demand; D ' F R ; D'FC; D'FT - same for the ex-post demand ( D ' F R * D FR ; D' F C * D FC ; D' F T *D' F T ); D p R . D ^ . D p r - same for the demand satisfied by transport mode mf, £)™j;D™2;D™3 - commercial characteristics describing the demand dependency on the supply quality, the load level and structure of the transport mode mi, respectively on the mode split and assignment for the transport system; D ™ / r - transport demand corresponding to transport mode mj, and route r.
Figure 2 The link between transport demand and transport flow Transport flow describes the use of the system's resources to satisfy transport demand. Therefore, it can be quantified using the number of loaded transport means and the distance. When the loads dispersion is significant a measure in gross tonnes kilometres is preferred. In any case, (vehicles kilometres or gross tonnes kilometres) the operation is reported to the calendar or operational time unit of the system (resulting in vehicles km/day or gross tonnes km/day for example). This expression, usually associated to the transport flow has the advantage of being easily summed and partially directly linked to the transport demand, and the disadvantage of not being able to characterise the transport flow. Hence, we believe that more complex characterisations of the transport flow are needed. These will benefit the evaluation techniques and policy making of the distribution logistics.
Transport demand, transport & traffic flow — key elements of city logistics 43 In this respect, the following can be identified: a) Physical characteristics, FTF: (i)
size and nature of transport flow, FT™, expressed through E.
, defined on the
(ii)
links(i, j) e D, of the modified graph (K, D) of the network I c K , D
b) Technological (functional) characteristics, FT-p: (i) homogenous/heterogeneous flow, throughout the trip (constant/variable load) (ii) equilibrium/variability per direction of travel (passenger different from freight transport) (iii) single mode, combined or multimodal transport (specially by using unit loads - pallets, containers, semi-trailers, etc - and solving the transhipment problem) (iv) quality parameters according to the logistic requirements (travel time, safety, compliance with delivery times) It is easily apparent that for the same demand characteristics different characteristics of the transport flow are obtained. This means that the same social demand can be satisfied with different resources. The transport technologies, adapted to the characteristics of the transport means and infrastructure networks determine the characteristics of the transport flow corresponding to a given transport demand. The evolution of the freight distribution systems offers a lot of conclusive examples in this respect. The direct distribution of the 50's developed into indirect distribution, with distribution centres and warehouses with (de)grouping functions, and is expected to further evolve into cross-docking. (Figure 3) (Binsbergen, 1998)
Figure 3 Logistics evolution
44
Logistics systems for sustainable cities
Of course the transport flows for some characteristics of the transport demand differ substantially. The use of unit loads, together with the evolution of the transport means and loading/unloading/storage systems, and also the development of information technology enabled the change of the distribution systems with major consequences for the transport flows and resources used. City logistics offers some other examples of mutations in the transport flow characteristics. (Tanaguchi et ah 2001; Visser and Binsbergen, 1998; Binsbergen and Visser, 1999; Boerkamps and Binsbergen, 1999)
TRAFFIC FLOW The superposition of the transport flows and other unit flows (trains, trucks, etc) or transport means (in empty state), and/or technological flows (for the maintenance needs of the transport system) over the infrastructure's elements (links, nodes) produces the traffic flows (Figure 4). As results from Figure 4 show, one has to differentiate between the transit (traffic) flows, corresponding to arcs D of the network N, and the input/output (traffic) flows, corresponding to the terminals and nodes K of the network N.
Figure 4 Formation of traffic flows In general, the transit flow is measured by the number of loaded/empty transport units and the number of technological transport units which use at one moment an element of the infrastructure, with the advantage of being easily summed and partially linked to the cost of using the infrastructure, and the transport system performances; and with the disadvantage of not having any direct link with the transport flow and not being able to completely characterise the traffic flow. For a more precise characterisation of the traffic flow, the following elements should be taken into account: a) Physical characteristics, FCF: (i) size, defined by a random variable with the appropriate probability density function
Transport demand, transport & traffic flow - key elements of city logistics 45 (ii)
(iii)
structure, usually heterogeneous, where from the necessity for some equivalence factor to be able to compare the mean traffic flow value with the transit or input/output capacity of the arc D or node K, and to identify the load factor synthetic parameters (mean flow speed, density, load factor, frequency, travel time, etc.)
b) Effects perceived by participants and residents, F C A (i) exogenous, environmental (air, water, soil, noise pollution, landscape and ecological equilibrium affected, etc) (ii) endogenous, internal (congestion, stress and accidents) A further analysis of the correlations between D, FT and FC reveals that the above mentioned physical characteristics of the traffic flow FCF, are just in part determined by the physical characteristics of the transport flow, and that the environmental and endogenous effects due to the existence of the flow could be the limiting cause for the transport need (a reduction in social mobility due to a transport supply impossible to further improve) — the dashed line in Figure 5. (where: M - transport means, N - infrastructure, and T - operation technology) The same figure shows that the traffic flow, FC, through the level of its characteristics and FCA) can inhibit or stimulate the transport flow development, FT, on that infrastructure element and can indirectly bring closer the ex-post demand to the ex-ante demand
CONCLUSIONS The ex-ante and ex-post transport demand, the transport and traffic flows are distinct notions, usually described by specific qualitative expressions, but these cannot completely characterise either of them. Each of these notions has to be described by physical, commercial, technological, environmental parameters, which allow precise and complete identifications. Between the analysed elements inverse links also exist (indicated in Figure 5 with wavy and dashed lines) which show that determining the ex-post demand D p , transport flow, FT and traffic flow, FC, using analytical models is a cumbersome and imprecise iterative process.
46
Logistics systems for sustainable cities
Figure 5 Complex dependencies between demand, transport and traffic flow The ex-post demand influences the value of the transport flow which in turn impacts upon the traffic flow value, but these links are not direct, and mainly because of the elements which differentiate the demand, transport means, technology and infrastructure, they cannot be completely described by analytical models. Hence, to evaluate the traffic flows and to predict their dynamics in order to establish the necessity and opportunity of infrastructure works is more indicated to use experimental determinations and empirical estimations. The quantitative measures proposed to describe transport demand, transport and traffic flow, and the diagrams which represent the links between them as well as the links with the elements of the transport system (infrastructures, transport means, technologies) in the authors' opinion bring the necessary details required for modelling the urban freight logistics. The outcomes of this research are useful to urban logistics both in the development of evaluation techniques and policy making. hi this respect, the objective of city logistics could be formulated as follows: for the transport demand with given characteristics, obeying the logistics constraints, the transport flow
Transport demand, transport & traffic flow - key elements of city logistics 47 (implicitly the resource use) and traffic flow (implicitly the resource use and environmental effects) have to be minimised. It is the task of theoretical and applied research to find ways of quantifying the freight distribution systems performance by identifying the quantitative links between the transport and traffic flow and satisfying certain social transport needs. At the same time, research has to remember the essential role of land-use, planning and restructuring of the socio-economical activities in order to modify those characteristics of the transport demand, which strongly influence the use of the transport system resources. Both research directions suggested represent synthetic expressions of the efforts to achieve sustainable transport. Due to its synthetic and mostly original character the paper is of course susceptible to additions and could generate criticism even though it represents the result of relatively extensive research preoccupations. We appreciate the input of those who by formulating opinions contribute to the necessary progress in the field of distribution logistics in general, and of city logistics in particular. The multidisciplinary, synthetic character of logistics imposes this sort of clarification, meant to reunite research activities of various specialists (engineers, mathematicians, statisticians, economists, urban planners, ecologists, etc). In this context, the paper would like to also constitute itself as a modest contribution to the field of logistics.
CASE STUDY This application proposes a system of indicators to perform connections between demand, transport flow and traffic flow as suggested before. It is the case of the distribution performed by a cosmetics company located in A which delivers the same quantity of goods (named hereafter unit loads or deliveries) to the retailers B, C, ...P. (Figure 6) The demands of the retailers are always for a unit load (box, carton) and occur randomly (Table 1, 2 present the situation of the week analysed). The supplier has three LCVs assigned to the retailers as shown in Figure 6. The deliveries have to be performed within a certain time window so they do not interfere with the trading hours. Thus, each LCV cannot make more than one trip daily (this restriction dictated the routing of the vehicles, too). The deliveries have to be just in time since the retailers have no storage facilities.
48
Logistics systems for sustainable cities
Figure 6 Location of supplier and retailers Two distribution strategies have been studied: (a) scenario 1 with delivery in maximum 24 h (Table 1); and (b) scenario 2 with delivery in maximum 48 h (Table 2) Table 1 Scenario 1: 24h delivery
In Table 1 and Table 2 are circled the demands satisfied by each vehicle in one trip. In Table 2 the trip is performed on the day corresponding to the lower part of the buffer. For example, the demands of B, E, F (day 2), and C, D (day 1) are performed by vehicle 1 on day2.
Transport demand, transport & trafficflow - key elements of city logistics 49 Table 2 Scenario 2: 48h delivery
Table 3 Routing for scenario 1
TOTAL WEEKLY DELIVERIES: 50
Table 4 Routing for scenario 2 VEHICLE 1 DAY
TOTAL TRIP EMPTY TRIP
ROUTE
A-J-G-H-I-L-A
1
VEHICLE 3
VEHICLE 2
TOTAL TRIP EMPTY TRIP
ROUTE
ROUTE
TOTAL TRIP EMPTY TRIP
27.8
n
A-B-C-E-F-D-A
m
A-B-C-D-E-A
9.6 29.5 12.4
A-J-G-H-l-A
13.5
45.1
6.0
15.6 53.1
23.6 10.0 A-K-O-M-N-A
29.9 6.9
V VI
vn WEEKLY TOTAL
A-C-D-E-F-A
31.5 11.8 28.6
A-B-C-E-A
12.4 24.4
A-B-F-A
28 .2
A-I-H-A
11 .8 25.0
A-I-G-A
A-P-M-N-A
12.2 A-K-N-A
11.8 B-4; D-3; C-4; E-4
F-3; 145.6 IS
58.0
1-4; G-3; L-1; J-2; H-3 13
104.6
EMPTY TRIP 8.5
A-O-P-A
IV
TOTAL TRIP 27.8
8.5 31.6
DAILY TOTAL:
K-2; M-2; P-2; O-Z; N-3
42.5
TOTAL WEEKLY DELIVERIES: 42
22.4 29.9 6.9 59.7
27.8
23.6 81.4
6.9 20.3
31.5 44.7
6.9
18.7
91.5
341.7
26.7
127.2
50 Logistics systems for sustainable cities Table 3 and Table 4 contain the summary of the vehicles activity and Table 5 the comparison between the two scenarios. Table 5 Comparison between scenarios
No
Indicator
1.
Traffic flow (total daily trip)
2.
Transport flow (daily loaded Irip)
3.
Average no of vehicles required daily
4. 5.
Total no of weekly deliveries (satisfied demand) Average no weekly deliveries by a vehicle
6.
Traffic flow per delivery
7.
Transport flow per delivery
8.
Transport flow per vehicle
9.
Traffic flow
10. Daily no deliveries by a vehicle Average no unit loads hauled by a 11. vehicle 12. Transport flow per vehicle
Units vehicles x km day loaded vehicles x km day vehicles day deliveries week deliveries vehicles x week vehicles km DO deliveries loaded vehicles x km no deliveries loaded vehicles x km vehicles loaded vehicles x km vehicles x km deliverieVday vehicles average no unit loads vehicle loadedvehiclesx (km/day)x avcragaio unitloads vehicles
Max delivery time Idav 2 days 64.8
48.8
36.9
30.6
3
1.86
50
42
16.6
22.5
9.1
8.1
5.2
5.1
12.2
16.5
0.57
0.63
2.4
3.2
1.7
2.1
20.7
34.6
One can easily observe that modifying the temporal characteristics of demand (24h, 48h service) has consequences upon the resources used by the carrier (see transport flow values), and also upon the load on the road infrastructure (see traffic flow values), (Table 5). Thus, the total transport flow decreases, the transport flow per delivery stays relatively constant, while the transport flow per vehicle, as a result of the reduction in the number of vehicles from 3 to 1.86 increases substantially - see row 8, and row 12 (where the average load of the loaded trip is taken into account). The better utilisation of the vehicle results from the comparison between the values of rows 10 and 11, too. The total traffic flow also decreases from 64.8 to 48.8 or from 9.1 to 8.1 on average for the service of a demand (see rows 1, 6). The same volume of demand was satisfied with a smaller number of deliveries (by grouping some demands with same destination in two successive days), row 4, which means that the favourable effects of resource consumption and of the external effects (including the infrastructure load) suppose modifications in the retailer's demand; by agreeing to a greater delay in delivery the retailer can enlarge his storage space and implicitly the stock, or he might accept a stock-out with the risk of losing some customers. The transport and traffic flow, in the various expressions proposed reflects: the efficiency of satisfying the transport demand (in the various logistic solutions of distribution, or when the
Transport demand, transport & traffic flow - key elements of city logistics
51
demand characteristics change); the consequences of modifying the demand characteristics when same distribution solutions are maintained. Therefore, we can appreciate that transport demand, transport and traffic flow, in these complex expressions proposed herein are key elements of the urban logistics. Of course, other interpretations are possible besides those presented in this case study.
REFERENCES Binsbergen, A.J. van (1998). New Logistics Concepts for Advanced Urban Freight Transport. Proceedings of European Transport Conference, Seminar B, P421, London, 93-104. Binsbergen, A.J. van and J.G.S.N. Visser (1999). New Urban Goods Distribution Systems. Urban Transport Systems Conference, Lund University. Boerkamps, J and A.J. van Binsbergen (1999). GoodTrip-A new Approach for Modelling and Evaluation of Urban Goods Distribution. Urban Transport Systems Conference, Lund University. Oppenheim, N. (1995). Urban Travel Demand Modelling. Wiley, New York Ortuzar, J de D, and L.G. Willumsen (1994). Modelling Transport- Second Edition, John Willey & Sons Ltd. Raicu, R (2002). On the modeling of sugar cane transport by road- PhD Thesis, School of Geoinformatics, Building and Planning, University of South Australia, Adelaide, pp. 98-139. Raicu, S. and M. Popa (1997). Cererea de transport - aspecte specifice. Revista Cdilor Ferate, 2, 19-26. Slinn, M., G. Peter and M. Paul (1998). Traffic Engineering Design- Arnold, London. Taniguchi, E., R.G. Thompson and T.Yamada (1999). Modelling city logistics. In: City Logistics I (E. Taniguchi and R.G. Thompson, eds.), Institute of Systems Science Research, Kyoto, 3-37. Taniguchi, E., R.G. Thompson, T. Yamada, and R. van Duin (2001). City Logistics. Network Modelling and Intelligent Transport Systems. Elsevier Science Ltd, UK. Visser, J.G.S.N. and A.J. van Binsbergen (1998). A Technology Push Towards Sustainable Urban Freight Transport. Proceedings of European Transport Conference, Seminar B, P421, London, 119-137.
This page is intentionally left blank
4
ASSESSMENT OF THE RELATIONSHIP BETWEEN VEHICLE TYPE MIX AND THE BENEFIT OF FREIGHT PROJECTS Kazuya Kawamura, University of Illinois, USA Anusha Seetharaman, Cambridge Systematics, Inc./Volpe Center, USA Saurav Dev Bhatta, University of Illinois, USA
ABSTRACT In the first part of this paper, a methodology for estimating the short-run economic impact of freight projects is proposed. The methodology quantifies three types of impacts; savings in travel time and distance, pavement damage, and broad economic development via multiplier effects. In the second part of the paper, the sensitivity of the impacts calculated by the proposed methodology with respect to the breakdown of vehicle miles travelled by vehicle configuration is assessed using a hypothetical road-widening project in Chicago. The analyses revealed that the overall economic impact is not particularly sensitive to the vehicle configuration since the decrease in pavement damage is only a small portion of the overall impact and also other costs are not affected significantly by vehicle configuration. In contrast, the split of the vehicle miles travelled between for-hire and in-house trucks is crucial since it has a direct effect on the value of time and subsequently, the cost reduction associated with travel time saving.
INTRODUCTION In the regional transportation planning arena in the United States, transportation of goods is often treated as an afterthought despite the fact that the current federal transportation funding authorisation bill, the Transportation Equity Act for the 21 s t Century (TEA-21) has recognised goods movement as one of the key planning goals (U.S. congress, 1998). While the efforts spearheaded by the Federal Highway Administration's (FHWA) Office of Freight Management and Operation, such as Freight Analysis Framework, have helped address some of the technical problems associated with forecasting and analysing freight movement, it is still apparent that
54 Logistics systems for sustainable cities the lack of project evaluation tools is one of the key reasons for the dire status of freight planning in the U.S. A survey by the FHWA revealed that while most (84%) of the metropolitan planning organisations1 (MPO) in the U.S. address freight planning as a part of the long range planning, only 45% have one or more freight projects in the implementation plan. Furthermore, it was found that only 13% of the MPOs possess freight-specific analytical tools. As a consequence, even though a high percentage of the MPOs list goods movement as one of the planning goals, a mere 10% use freight-specific criteria during the project evaluation and selection processes (FHWA, 2002). Fortunately, in the last several years, some MPOs and states have made efforts to integrate freight transport into the long-range planning process. However, in most cases, goods movement forecasts were made by an ad-hoc adoption of passenger demand forecasting models. Typically, heavy vehicle factors, based on VMT estimates or the registered number of heavy vehicles, are used to estimate the truck volumes based on the passenger vehicle volumes. The accuracy and sensitivity attained by such techniques are obviously not adequate for evaluating the costs and benefits of proposed freight projects. Another important issue that has a serious and complex implication for goods movement planning is "environmental justice". The Presidential Executive Order 12898 (The President of the United States, 1994), and the Department of Transportation (DOT) Final Order on Environmental Justice in Minority Populations and Low-Income Populations (USDOT, 1997) require MPOs to address the distribution of costs and benefits associated with proposed transportation projects. Thus, it is critical to identify, as a part of goods movement planning, the impact on the low-income population who often reside near terminals or along freight corridors. hi light of the backdrop discussed above, this paper strives to develop a methodology and identify the data requirements for evaluating the impact of freight corridor projects. Specifically, in the following sections, key types of impacts associated with truck traffic are identified. Then a collection of methodologies to estimate the magnitude of the impact are introduced. Finally, a case study, using hypothetical widening of Cicero Avenue, which is one of the major truck routes in the City of Chicago, is conducted to assess the effect that different mix of heavy vehicle types have on the estimated magnitude of impacts. The goal of the study is to identify the level of detail, in terms of vehicle types, the travel demand forecasting model must provide in order to assess the costs and benefits associated with freight projects with reasonable accuracy.
PROJECT Cicero Avenue is one of the major freight corridors in the City of Chicago. The land use along the study segment is mostly industrial. For the last few decades, most of the economic growths in the region have taken place either in the downtown area, commonly known as the "Loop", or in the suburbs. In particular, the area around the O'Hare Airport has recently become the largest employment node in the region, surpassing the Loop. 1
MPOs are responsible for the long and short-range transportation planning at regional level.
Benefit for freight projects
55
This economic shift and also the migration of the population to the suburbs have left older industrial corridors such as Cicero Avenue in a precarious position. Most of the major infrastructure projects that are programmed into the long-range plan for the region are in the suburbs. Thus, it is unlikely that Cicero Avenue and other older streets that frustrate truck drivers with height restrictions (at overpasses), poor pavement conditions, and congestion will be improved in the near future. As nearly all major cities in the U.S. have experienced similar or even more pronounced suburbanisation during the last few decades, the situation in Chicago is not unique. The motivation for selecting such facility for this study is to find out whether improving industrial corridors that may not attract the attentions of the commuters and decision makers but are still vital to the flow of goods would bring a benefit to the entire region. Cicero Avenue is a principal arterial with a 4 or 6 lane cross-section, and carries approximately 78,000 AADT. Heavy vehicle traffic accounts for 22% of the AADT, but during the peak period for truck traffic, which is between 12:30 PM and 1:30 PM, over 36% of the traffic are trucks2. The hypothetical project that is the focus of this study widens an approximately 10 km stretch of Cicero Avenue, between Grand Avenue to the north and Ogden Avenue to the south, from 4 lanes to 6 lanes. This would make Cicero Avenue a 6-lane arterial all the way from the freeway to the north, I-90/I-94, to 1-55 in the south. The project area and its spatial relationship with the major facilities and freight terminals in the region are depicted in Figure 1. The map reveals the importance of Cicero Avenue in serving the freight flow in the region. Firstly, it can be used by the trucks to traverse between three major expressways, which run primarily eastto-west, without going to the Central Business District. Also, Cicero Avenue is used by the trucks that transport goods between the two major airports, O'Hare and Midway, and numerous intermodal terminals and the factories in the industrial sections of the city.
IMPACT ANALYSIS METHODOLOGIES Kawamura and Seetharaman (2002) developed a comprehensive evaluation framework for freight projects based on the goals set by TEA-21. The framework consists of six modules that calculate the impacts associated with: 1) travel time saving, 2) accident, 3) noise pollution, 4) air pollution, 5) pavement damage, and 6) economic development. In this study three types of impacts; savings in travel time and distance, pavement damage, and economic development, are calculated. It should be noted that in this study, the costs associated with environmental impacts, including air and noise pollution, and accidents were not included. Monetising those impacts will require valuations of quality of life, health conditions, and even human lives. While much advancement has been made in the fields of public health and medicine to conduct such valuation, to our knowledge, no consensus has been reached on the methodology or appropriate dollar values for those types of impacts. Thus, it is our hope that in the long run, further research efforts will be conducted to collect information that will enable us to incorporate those impacts in the framework presented in this paper. Following is a description of the methodologies that are used to calculate those three types of impacts for a hypothetical widening of Cicero Avenue as discussed in the previous section.
1
Figures are based on the traffic counts conducted by the City of Chicago
56 Logistics systems for sustainable cities Travel time and distance savings Travel time savings associated with the proposed project can be monetised using value of time. Since the framework calculates the saving in vehicle operating cost within the economic impact analysis, only labour and fringe costs should be included in the calculation of value of time. Equation 1 gives a simple way to calculate the monetary value of travel time on a network from demand forecasting model outputs. Value of Truck Travel Time ^ ^ x g x ^ x C , ) (1) i
k
where, TTj = Travel time on link i from a model Qj = Truck volume on link i Pjk = Proportion of truck type k on link i Ci, = Mean value of time for truck type k. Truck travel demandforecasting The truck volumes and travel times were estimated using the traffic assignment module of the EMME2 traffic demand forecasting program. The parameters and input data were taken from the model developed by the Chicago Area Transportation Study, the MPO for the Chicago region. The CATS model follows the standard Urban Transportation Planning System, commonly known as the "four-step" process.
Benefit for freight projects
57
The model divides a day into eight time periods. It forecasts travel demand for each of the eight time periods in a sequential manner using the final link travel times from the previous time periods as the link speeds for the first iteration of the equilibrium assignment process. For commercial vehicle trips, the model generates separate trip tables for four distinct vehicle types; commercial automobiles, light trucks (under 12,700 kilograms in gross vehicle weight), medium trucks (between 12,701 and 29,000 kilograms), and heavy trucks (between 29,001 and 36,000 kilograms). Due to the increase in the number of Sports Utility Vehicles that are classified as light trucks, but used for non-commercial (i.e. household travel) purposes 3 , for this study, the costs and benefits associated with only the latter two vehicle types were analysed. Also, only the first six time periods, covering from 8PM to 4 PM, were analysed since only a very small portion of the medium and heavy truck trips occur during the PM peak periods. The values of C^ were taken from a study by Kawamura (2000). Kawamura's estimate of the value of time was derived from a stated preference survey of truck operators in California. While the studies by others, including Jack Faucett and Associates (1990), and FHWA (2000), found that values of time are generally not associated with vehicle type and weight, Kawamura !
In the model, these are included in the passenger trip, not the light truck category.
58 Logistics systems for sustainable cities found that operator types (i.e. for-hire versus in-house) have a considerable effect. However, the framework of the demand forecasting model used for this study does not distinguish the truck trips by operator types. Therefore, the weighted average of the values of travel time was estimated using the VMTs for both types of trucks, obtained from the Vehicle Inventory and Use Survey (VIUS) as the weights. Table 1 shows the average values of time, taken from Kawamura's study, and the VMT shares for for-hire and in-house trucks that were used to calculate the weighted average. Table 1 Value of Time and VMT Shares Average value of time ($/hr.) VMT share 88.6% For-hire trucks 28.1 11.4% In-house trucks 17.6 (Kawamura, 2000) In the U.S., the public disclosure of the data concerning the operation of trucks is strictly controlled in order to protect the interest of truck owners that participate in the surveys. Thus, it is always a challenge to extract relevant information from the data sources such as VIUS. VIUS is a national survey of truck owners, including both for-hire and in-house fleets, conducted every 5 years to obtain the information about usage and vehicle characteristics of commercial vehicles. In 1997, VIUS sampled about 131,000 out of a total of 75 million trucks in the U.S., an extraction rate of about 1.7%. The truck owners selected for the survey are obligated by law to respond, resulting in a high response rate of about 85%. For calculating the VMT share, the records in the VIUS that belong to the trucks with gross vehicle weight (GVW) of over 13,000 kg and based in one of the metropolitan statistical areas in the state of Illinois were first extracted. Then, the trucks that operate as long-range carriers (over 320 kilometres) were eliminated because they were unlikely to operate in the study area. While equation 1 captures the costs or benefits that accrue with a change in travel time, it does not directly measure the effects of distance related gains and losses such as depreciation, fuel, vehicle maintenance and repair, and tyre costs. Since the improvement in Cicero Avenue will affect the travel behaviour for the entire transportation network for the region, the impacts associated with the decrease or increase in the travel distance must be assessed. The figures in Table 2 suggest that distance-related costs do not vary significantly with the number of axles. However, there are more noticeable variations with respect to the GVW. Thus, the cost components for the most common truck configuration, truckload (TL) 5-axle 48' vehicles, were used in the calculations. After being adjusted to the 2000 dollars, the cost components for the 23,800 kg GVW and 35,400 kg GVW were multiplied by the VMTs for medium trucks and heavy trucks from the demand forecasting model, respectively, to obtain the estimate of the distance-related truck operating costs.
Benefit for freight projects
59
Table 2 Distance-Related Truck Operating Cost Cost Component (1988 cents per km) Size Vehicle Fuel Tires Repair Total GVW (1000 kg) 3.8 28.6 10.4 1.9 12.5 TL 5 axle 48' 12.7 5.3 31.6 11.9 1.9 12.5 23.8 6.8 35.0 13.5 2.2 12.5 35.4 3.9 29.9 10.5 2.6 13.0 TL 6 axle 48' 13.4 5.4 32.9 12.0 2.6 13.0 24.5 6.9 36.2 13.6 2.6 13.0 36.1 4.4 30.6 11.3 1.9 13.0 14.2 TL 5 axle twin 28' 6.1 34.1 13.1 1.9 13.0 27.1 1.1 30.8 14.4 2.3 13.0 36.3 5.4 29.4 12.1 1.9 10.0 25.2 LTL 5 axle 48' 3.8 27.9 10.1 1.9 12.3 Flat beds 5 axle 48' 12.5 6.8 34.9 13.6 2.2 12.3 35.4 3.6 1.9 9.7 33.2 18.0 11.2 Tank 5 axle 42' 6.8 40.4 13.4 2.2 18.0 35.4 Source: FHWA, 2000 Vehicle Type
Cost of pavement damage Another major impact that mainly depends on VMT is the cost of pavement. Unlike vehicle operating costs, the cost of the damage to the pavement caused by truck traffic is borne by the public. The best unit of measurement of pavement damage is the cost to repair the damage caused by truck traffic. Recent studies such as the Truck Size and Weight Study (FHWA, 2000) and Highway Cost Allocation Study (HCAS) (FHWA, 1997a) provide excellent information for determining the impact on pavement and bridges. Highway Cost Allocation Study is conducted by the Federal Highway Administration for the purpose of analysing highway cost responsibilities for different vehicle classes by the federal and state transportation agencies. In the simplest terms, the study allocates highway and bridge expenditures, including construction, repair, right-of-way, provision of rest areas and climbing lanes, administration and overhead, and planning, to 12 vehicle classes based on vehicle miles travelled (VMT) and impacts. The expenditures are estimated from the financial records of the agencies. The VMT for each vehicle type on various roadway classes and pavement types were estimated from a combination of data sources including Highway Performance Monitoring System (HPMS), Vehicle Inventory and Use Survey (VIUS), and Census of Transportation. In order to estimate the shares of pavement damage, a new pavement deterioration model, named the National Pavement Cost Model (NAPCOM), was used in conjunction with the equivalent single axle load (EASL). Selected results from the Truck Size and Weight Study, which used the HCAS to calculate the impacts of truck regulations, arc shown in Table 3. Again, the changes in the total VMT for medium and heavy trucks resulting from the widening of Cicero Avenue were estimated by the travel demand forecasting model. Since the pavement impact differs considerably by the type of road and also the type of vehicles, the medium and heavy truck VMTs for each of the six types of roads had to be estimated.
60 Logistics systems for sustainable cities Table 3 Pavement Impact in Dollars per km (in 1994 dollars) Single Unit Semi-trailer 4-axlcs Vehicle Configuration 3-axles 5-axles 6-axles GVW(kg) 24,500 29,050 36,320 40,860 Urban Interstate 0.04 0.03 0.03 0.03 0.06 0.04 0.04 0.03 Urban Expressway 0.08 0.08 0.06 0.06 Urban Principal Arterial Urban Minor Arterial 0.19 0.15 0.14 0.11 Urban Collector 0.41 0.44 0.34 0.31 Urban Local 1.46 1.58 1.19 1.09 Source: FHWA, 2000 Also, since the model does not produce VMT estimates for specific vehicle configurations, it was necessary to estimate appropriate average costs for medium and heavy trucks. The VIUS data again were used for that purpose. The VIUS data indicate that over 85% of the VMT for the heavy trucks (i.e. over 29,001 kg in GVW) are attributed to semi-trailers with 5-axle configuration. Thus, the unit pavement costs for that vehicle type were used to calculate the pavement costs associated with the heavy truck travel. For medium trucks (i.e. between 12,701 kg and 29,000 kg), there is no single dominant vehicle configuration. Thus, the VMTs for the two most popular vehicle types in the medium truck category, 3-axle single unit and 5-axle semi-trailer, extracted from the VIUS dataset were used to calculate the weighted average of the unit pavement costs. Table 4 shows the unit costs used for the calculation of pavement maintenance and repair costs. Table 4 Unit Pavement Cost Used (in 2000 dollars per kilometre) Medium Trucks Heavy Trucks 0.03 0.04 Urban Interstate 0.05 0.06 Urban Expressway 0.07 Urban Principal Arterial 0.08 0.16 0.18 Urban Minor Arterial 0.38 0.42 Urban Collector 1.48 1.34 Urban Local Economic impact analysis The Input-output model is a popular tool for quantifying broad economic impacts of major projects such as ports, shopping centres, sports facilities, and roads. In the early 1980's, inputoutput models were developed under the sponsorship of NCHRP for the purpose of evaluating economic impacts of transportation projects (TRB, 1995). However, these models were developed to assess the economic impacts associated with construction activities and did not have a capability to reflect increased productivity due to improvement in the movement of freight. To address this shortcoming, Seetharaman et al. (2003) used the price elasticity of demand for transportation and warehousing sector to estimate the increase in output associated with the expected reduction in congestion (and thus truck operating cost) from the collection of projects included in the $40 billion regional transportation plan (CATS, 1998) for the Chicago region. The method first calculates the percent reduction in the cost of running trucks based on the travel time and fuel cost savings. Then, price elasticity is applied, under the assumption that the competitive nature of the trucking industry in the Chicago region will force the companies to lower the rate, to estimate the change in the demand for truck shipments. Under the
Benefit for freight projects
61
assumption that increase in demand will be met by the rise in supply, estimated increase in output is fed into the input-output model, IMPLAN, to calculate the overall economic impact. Input-output analysis uses the data obtained from the financial transactions to quantify the inter-dependence among industries. If there is an increase in the goods or services provided by an industry sector, its impact permeates the entire economy through those inter-industry relationships. In input-output terminology, the former is called "direct" impact while the latter is known as "indirect" impacts. Since the first round of indirect impact will generate additional inter-industry transactions, multiple rounds of simulations are conducted until the impact becomes insignificant through various "leakages" such as international or inter-regional transactions, savings, and taxes. The third and final type of impact analysed in the input-output framework is called "induced" effects. The induced impact is generated by the changes in wages and salaries, and ultimately household expenditure, which in turn initiates another round of inter-industry transactions. This study utilised the input-output data generated for the 528 industry sectors in the Chicago Region for the aforementioned study by Seetharaman et al. The approach used in this study to identify the initial change in the freight sector output is similar to the one employed by Seetharaman et al with the exception of several modifications. First, in addition to the travel time saving, the reduction in travel distance was obtained from the demand forecasting model. Then, the savings in user costs, including both travel time and distance but excluding pavement damage, were used to estimate the reduction in truck operating costs. Once the cost reduction was obtained, price-elasticity of -1.0, a figure estimated by Friedlaender and Spady (1980) was used to calculate the change in the total output of the trucking and warehousing sector of the economy.
RESULTS Travel time and distance savings The travel times from the demand forecasting model runs for with and without the Cicero Avenue improvement are shown in Table 5.
Automobile Medium truck Heavy Truck
Table 5 Network Travel Times Before Improvement After Improvement 130,792,782 130,738,317 212,177 210,670 6,056,141 6,053,099
Difference -54,465 (-0.042%) -1,506 (-0.7105%) -3,042 (-0.050%)
The figures show that in terms of percentages, expected reduction in travel time is very small for all three vehicle types shown. As mentioned previously, the average value of time, weighted by the for-hire and in-house truck VMTs obtained form VIUS, was applied to the time saving to estimate the monetary value of travel time reduction. Despite the small percentage reduction, it was estimated that over $122,000 worth of travel time reduction per day could be expected from the widening of Cicero Avenue. The changes in travel distances, estimated by the demand forecasting model, are shown in Table 6. Interestingly, the model forecasts did not show any change in the distance travelled by
62 Logistics systems for sustainable cities medium trucks, indicating there would be no route change to take place in response to the project. Also, the widening of Cicero Avenue would have greater impact on the heavy truck travel than automobiles. Furthermore, for heavy trucks, it will increase the travel on the expressways while reducing the travel on major and principal arterials.
Automobile Medium truck Heavy Truck
Prin. Art. 587 (0.001%)
Table 6 Change Maj. Art -61,841 (-0.261%)
0 -33,220 (-0.809%)
0 -216,191 (-9.396%)
in Network Miles Travelled Min. Art. Collector Expressway -37,631 -4,539 7,520 (-0.372%) (-0.162%) (0.021%)
Total -95,904 (-0.083%)
0 1,000 (0.153%)
0 -213,074 (-1.047%)
0 458 (0.303%)
0 34,879 (0.266%)
Expected reductions in vehicle operating costs, categorised into four expense types, are shown in Table 7. Among four types of expenses, most substantial savings are in fuel cost and vehicle depreciation. The total amount of saving is about $94,000 per day. Thus, the total reduction in use costs, which include both time and distance related costs, would be over $216,000 per day. Table 7 Reduction in Distance-Related Operating Fuel Tires 0 Medium truck 0 -36,500 -5,900 Heavy Truck -36,500 -5,900 Total
Costs (per day in 2000 dollars) Repair Vehicle 0 0 -18,400 -33,800 -18,400 -33,800
Cost of pavement damage The unit pavement costs, shown in Table 3, were applied to the change in the heavy track VMT for each type of roadway to estimate the cost impact. The cost reduction is mainly generated by the reduction in total miles driven as well as diversion of traffic from principal and major arterials to expressways, which have lower unit pavement cost. Table 8 shows the change in the cost of pavement damage for each road type. Table 8 Change in the Cost of Pavement Damage Prin. Art. Maj. Art Min. Art. Collector Expressway Change in -2,300 Costs
-15,100
100
200
1,000
Total -16,100
Breakdown of cost reductions Table 9 and Figure 2 show the breakdown of the cost reductions. The cost reduction from travel time saving, which mostly accrues from labour cost saving, contributes the greatest share. This is not surprising since the most significant contribution of a lane addition, such as the one studied here, would be a decrease in travel time. However, it is surprising that only 7% of the cost reduction is attributed to the pavement damage. This is probably due to the fact that a very small percentage of truck travel miles are on minor arterial and collector roads, which are much more susceptible to damage by track traffic, while most of the reduction in the track VMT occurred mainly for the larger surface streets and expressways. The total amount of saving,
Benefit for freight projects
63
$233,100 per day, is a considerable amount since it can translate into over $50 million dollars of saving per year.
Table 9 Aggregate Cost Reduction per Day (in 2000 dollars) Pavemen t Fuel Tires Repair Vehicle Time N/A 0 0 0 0 Medium truck 0 -16,100 Heavy Truck -36,500 -5,900 -18,400 -33,800 N/A -36,500 -5,900 -18,400 -33,800 -122,400 -16,100 Total
Total N/A N/A
-233,100
Economic impact Despite the popular image of Chicago as the largest freight hub in the U.S., the figures in Table 10 show that commercial freight and warehousing sector only rakes 20th in the regional economy in economic output4. Nevertheless, the total economic output of the said sector is over $7 billion a year.
4
As stated previously, this assumes all the cost saving will go into the reduction of price the trucking businesses charge.
64 Logistics systems for sustainable cities
Rank 1 2 3 4 5
20
Table 10 Annual Production and Employment in the Chicago Region Industry Output Employment (millions 1997$) Real estate 52,899 119,779 Wholesale Trade 37,734 278,953 30,007 Retail Trade 718,617 Business services 26,751 415,485 Construction 24,117 242,195
Motor Carrier Warehousing
Freight
Total (all sectors)
and
7,878
499,615
67,189
4,768,035 (Seetharaman et al., 2002)
As shown in Table 11, the operating cost reduction translates to about $10 million worth of growth in freight sector output. This increase in freight sector output is expected to generate an additional increase of $3.3 million in economic output through inter-industry transactions, which is called "indirect" impact. Table 11 Economic Impact of Freight Sector Output Indirect Direct Sector & Motor Freight 2,424,000 10,383,000 Warehousing 227,000 0 Real Estate 486,000 0 Wholesale Trade 92,000 0 Retail Trade 0 102,000 Construction
Total (all sectors)
10,383,000
3,331,000
Increase (in 2000 dollars) Total Induced 136,000 934,000 636,000 1,013,000 792,000
12,943,000 1,161,000 1,122,000 1,105,000 894,000
3,511,000
17,225,000
Finally, the model estimated that about $3.5 million of economic growth would be generated by the "induced" effect, which mainly consists of an increase in household expenditure and government revenue. Therefore, it is estimated that a total of about $17 million dollars worth of economic growth is generated by the project. It should be noted that this figure does not include the impact of the money spent on the actual construction. Furthermore, this should be regarded as a long-run economic impact because the assumptions employed in the analysis, especially the premise that the reduction in operating costs would eventually lead to an increase
Benefit for freight projects
65
in the freight sector output, make this type of impact mutually exclusive from the user cost reductions identified in the previous section5.
SUMMARY In this study, a practical approach that can be applied within the current transportation planning process is used to assess the impacts of a capacity enhancement project of a corridor that is used heavily by trucks. In the calculation of the benefits, a considerable effort was made to incorporate the effects of vehicle configuration and weight rather than treating trucks as a homogeneous group of travellers as commonly done in planning studies. Analysis results suggest that the assumptions regarding the vehicle configuration do not have a significant effect on the overall cost savings for the type of project analysed. This is because the decrease in the damage to the pavement is only a small portion of the overall cost reduction and also the user cost does not vary substantially with respect to vehicle configuration. Furthermore, this particular project involved an improvement to the arterial that is a major feeding route to the expressways. Thus, the project diverted truck traffic to the expressways, with a durable pavement design, rather than to other surface streets. It also should be noted that construction of a new route, as opposed to the lane addition, will require further study since such project may have far greater effect on the network travel distance, and subsequently results in a much greater impact on the pavement damage. In contrast, the assumption regarding the shares of truck VMT by for-hire and in-house trucks was crucial since it has a direct effect on the value of time and subsequently, the cost reduction associated with travel time saving. Unfortunately, the only source of such data available in the U.S. is the VIUS, which is publicly released only at an aggregate level. The total impact of the project is considerable. As noted previously, user cost saving alone can exceed $50 million a year. Even though the project would be likely to cost over $150 million, the benefit to the freight industry alone can recover the cost in several years. In the long-run, the user cost saving will be replaced by the increased output of the freight sector, which is expected to expand the regional economy by approximately $17million. Unfortunately, in regional planning process in the U.S., the benefits to the freight industry are usually overlooked when funding priorities are set. As a result, the projects such as the one studied in this paper do not receive adequate attention from the planners and decision makers. It is our hope that the methods presented in this paper may help planners to develop an evaluation framework that can be used to assess the potential benefit of freight-oriented projects.
REFERENCES Chicago Area Transportation Study, (1998). 2020 Regional Transportation Plan- Chicago Area Transportation Study.
5
The user cost analysis assumes the number of trip ends to remain constant. Meanwhile, the increase in freight sector output is achievable only through an increase in the number of revenue-generating trips.
66 Logistics systems for sustainable cities Federal Highway Administration, Office of Freight Management and Operation (Accessed 1020-2002). Freight Survey Analysis. http://www. ops.fhwa.dot.gov/freigh t/freight_planning.htm Federal Highway Administration, (2000). U.S. DOT Comprehensive Truck Size and Weight Study. Federal Highway Administration, Washington D.C. Federal Highway Administration, (1997a). Federal Highway Cost Allocation Study: Final Report, U.S. Department of Transportation, Washington D.C. Federal Highway Administration, (1997b). 23 CFR 772, "Procedures for Abatement of Highway Traffic Noise and Construction Noise". Federal Highway Administration. Washington D.C. Friedlaender, A. and R. Spady, (1980). A Derived Demand Function for Freight Transportation, Review of Economics and Statistics, 62,432-441. Jack Faucett Associates (1990), The Effects of Size and Weight Limits on Truck Costs, Working Paper, prepared for the Federal Highway Administration. Kawamura, K. (2000). Perceived Value of Time for Truck Operators, Transportation Research Record 1725, Transportation Research Board, Washington, D.C. Kawamura, K. and A. Seetharaman (2002). Data Requirements for Truck Project Impact Evaluation. Proceedings from the 2002 Transportation Research Board Annual Meeting. Transportation Research Board. Washington D.C. Seetharaman, A., K. Kawamura, and S. D. Bhatta (2003). "Economic Benefits of Freight Policy relating to the Trucking Industry: An Evaluation of the RTP Freight Policy for the Chicago Six County Region: forthcoming in Transportation Research Record. Transportation Research Board, (1995). NCHRP 8-30. National Academy Press, Washington D.C. The President of the United States (1994). Executive Order 12898, Federal Actions to Address Environmental Justice in Minority Populations and Low Income Populations. Federal Register 59:32, 7629-7633. U.S. Congress, (1998). The Transportation Equity Act for the 21st Century: Washington D.C, June 1998. U.S. Department of Transportation, (1997). Final Order on Environmental Justice, Order 5610.2, Federal Register 62:72, 18377-18381.
5
ESTIMATION OF AN ORIGIN-DESTINATION MATRIX FOR URBAN FREIGHT TRANSPORT. APPLICATION TO THE CITY OF SEVILLE
Jesus Munuzuri, School of Engineering, University of Seville, Spain Juan Larraneta, School of Engineering, University of Seville, Spain Luis Onieva, School of Engineering, University of Seville, Spain Pablo Cortes, School of Engineering, University of Seville, Spain
ABSTRACT This work presents the process followed to estimate an origin-destination matrix for freight transport in the city of Seville. The focus lies specially on the data acquisition process, which is normally one of the main obstacles to be faced by this type of analysis. The model used, based on entropy maximisation, and the solution algorithm, including Frank-Wolfe's linear approximations, are equally described. Results are presented graphically and then compared with vehicle flows observed from reality, allowing to extract conclusions for the validation of the model.
INTRODUCTION This paper describes work that is part of a project aimed at obtaining a description of freight movements in the city of Seville. The main tool for carrying out the analysis is the commercial package EMME/2, which performs equilibrium assignments (traffic flow calculations) using an urban network divided into zones and an origin-destination (O-D) matrix, which reflects how many vehicles travel between every pair of zones. An O-D matrix for passenger automobile traffic was already available for the city, and so passenger traffic assignments were already performed. This paper will show the process followed for estimating O-D matrices for freight transport vehicles, as well as carrying out a passenger automobile traffic, freight traffic
68 Logistics systems for sustainable cities conjoint assignment. A comparison between estimated and actual freight flows will then be made, identifying the weaknesses of the model and the need for further information. The estimated O-D freight vehicle matrices, as well as the one for passenger automobile traffic, correspond to the interval between 9 and 10 in the morning of a normal workday. It is important to note that, whereas private vehicles usually follow a single purpose movement (i.e., from home to work), freight vehicles follow a route with many movements and stops in order to deliver to different parts of the city. However, the assumption was made here that in the time interval considered, freight vehicles go towards their first stop in the route, and this is the mobility that will be considered here. This assumption implies that, once the vehicle has arrived to a destination zone, all the delivery points for its route are located inside that zone.
SOURCES OF INFORMATION The lack of valid information is a permanent burden upon freight transportation model building. According to Cohen (1997), two types of information sources, primary and secondary, can be used when building a model. Primary sources (such as vehicle counts and logistic practice patterns) provide much better information for building the model, but are much more difficult and costly to access than secondary ones (such as aggregate statistics, and census data). An adequate combination of the use of both types of sources is therefore required for improving the modelling process.
Secondary sources of information The secondary sources of information used in this case were: (a) Population densities, obtained from the Seville Demographic Office. The number of residents in each zone of the city are shown in Table 1 (b) Business locations, obtained from the Seville Chamber of Commerce. This includes the number of wholesalers and retailers located in each zone of the city according to the type of activity. Due to its extension, this information will not be displayed here, but it can be found in full extent in Munuzuri (2003) (c) Activity sectors, extracted from classification of the previous information. The five sectors considered in the analysis, which contain all types of commercial activities which imply freight movement in Seville, are: 1. Transformation of mineral non-metallic products and chemicals: construction, pharmacies, glass products, paints, washing products, etc. 2. Transformation of metals: machinery, computers, electric equipment, vehicles, etc. 3. Other industries: textile products, shoes, furniture, paper products, financial institutions, etc. 4. Non-fresh food products, beverages and tobacco. 5. Fresh food products: fruit, vegetables, meat and fish. (d) Number of vans and trucks with license for freight transportation issued in Seville, provided by the Andalusian Federation for Road Freight Transport (FATRANS). This number equals to 5,823 and will be used as the total number of freight vehicles operating in Seville. Of course, every day there is a number of freight vehicles operating in Seville with licenses issued elsewhere in Spain, but this was assumed to be equivalent to the number of vehicles with a license issued in Seville and operating somewhere else.
Estimation of an O-D matrix for freight transport
69
4779 8859 9119 3896 7642 6885 8559 4481 3839 6944
644 9920 5675 8885 1504 5751 1038 3204 6857 7582 4197 2619 8289 3320 6029 4007 4025 5241 5832 6287
67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99
41006 41006 41006 41006 41006 41006 41006 41006 41013 41013 41013 41013 41013 41013 41013 41005 41013 41013 41005 41013 41013 41013 41013 41013 41016 41012 41012 41012 41011 41011 41011 41011 41011
9227 9355 7946 11284 2333 11914 13380 4767
627 7086 4973 2025 1100 6857 2407 5146 8430 3556 2025 5457 8391 9499 7799
464 1084 2988 5969 2247
0 4319 52323 5920 9599
POPULATION
860 860
POSTAL CODE
9118
ZONE
POPULATION
41008 41008 41008 41008 41008 41008 41003 41003 41008 41008 41008 41007 41007 41007 41007 41007 41007 41007 41007 41004 41005 41005 41005 41005 41005 41005 41005 41005 41005 41005 41006 41006 41006
POPULATION
POSTAL CODE
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66
ZONE
ZONE
41001 3577 41001 2679 41004 3137 41003 2572 41004 2566 41004 2146 41001 3794 41004 1011 41002 5959 41002 5588 41003 7404 41003 4069 41003 6404 41003 4178 41009 6447 41009 920 41009 4758 41009 393 41009 4308 41015 5828 41009 2680 41015 786 41009 7084 41008 3640 41008 6518 41009 6229 41008 9586 41008 9518 41008 7898 41009 12440 41009 12961 41008 11890 41008 4780
POSTAL CODE
POPULATION
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
POSTAL CODE
ZONE
Table 1 Population in the different zones of the city.
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129
41011 41011 41011 41010 41010 41010 41010 41010 41010 41010 41092 41092 41092 41020 41020 41020 41007 41020 41019 41020 41020 41020 41017 41016 41017 41016 41016 41016 41014 41016
707 936 909 4746 7314 8074 8729 7793 6028 9059
0 1150
0 1450 2968 13132
676 13656 8620 9034
700 1300 3000 26541 13105
845 973 6548 11604
250
70
Logistics systems for sustainable cities
1 Pharmacies Perfumes, druggists Plants Petrol stations Seeds Waste disposal premises Urban cleaning Cleaning services Dry cleaning of clothes Construction materials 2 Home appliances Domestic gear and hardware Doors, windows and blinds •lome equipment Vehicles Spare parts for vehicles Boats and ships Machinery Tyres Office appliances Vledical gear Other machinery Technical services for home appliances Mechanics for vehicles Technical services for consumer goods Mechanics for industrial machinery Other technical services 3 Confectioneries and pastries Bakeries [ce creams and sweets Fried food Tobacco
1 0.2 0.2 1 1 0.2 0.2 0.2 0.2 1 0.4 1 1 1 0.2 1 0.2 0.4 1 0.4 0.4 1 1 0.4 0.4 0.4 0.4 2 2 1 1 1
3 Food and beverages Food shops Food shops without premises Food and beverages on machines Department stores Restaurants Coffee and chocolate Bars and pubs Food and beverages without premises Other foods and beverages 4 Hotels, motels and boarding houses Home textiles Clothes Underwear Shoes Leather and furs Furniture (except office) Second-hand goods and furniture Music instruments and accessories Stamps, coins and medals Books, newspapers and magazines Jewellers and watch shops Toys, weapons and sport items Other goods Textiles and clothes without premises Shoes without premises Other goods without premises 5 Fruits and vegetables Butcher's shops Eggs and farm products Fish
DELIVERIES PER DAY
PREMISES INCLUDED
SECTOR
DELIVERIES PER DAY
PREMISES INCLUDED
SECTOR
Table 2 Average estimated number of deliveries per day for all types of premises considered within the five sectors.
3 11 1 1 2 2 2 2 1 2 1 0.4 0.4 0.4 0.4 0.4 2 1 0.6 0.4 3 0.4 0.6 1 0.2 0.2 0.2 1 1 1 1
Primary sources of information The primary sources of information used were: (a) Freight vehicle counts in different streets of the city, used for model validation. (b) Surveys carried out among a significant number of retailers in different parts of the city (Munuzuri, 2003). These surveys provided two different types of data: (i) Average number of deliveries received daily by each type of retailer. This information is shown in Table 2.
Estimation of an O-D matrix for freight transport (ii)
71
Average number of vehicles used for home deliveries by each type of retailer. It was found out that mainly four types of retailers deliver their goods to their final customers: furniture retailers (sector 4), home appliances (sector 2), office appliances (sector 2), and doors, windows and blinds (sector 2). It was also determined that the average number of vehicles used for home deliveries is one per retailer, and so the estimated number of vehicles is then 1,028. This number corresponds to the total amount of the mentioned premises located throughout the city of Seville, which was extracted from the business location information, presented in Mufiuzuri (2003). Of course, some retailers employ more than one vehicle for home deliveries, but then again many of them use the same vehicle for several stores.
THE ENTHROPY MAXIMISATION MODEL The objective of the estimation process was to calculate the values of Ty, that is, the number of freight transport trips between zone / and zoney, for all (ij) zone pairs in the city. These Ttj values are the elements of the freight Origin-Destination matrix. The analytical tool used for the estimation of O-D freight vehicle matrices was the Enthropy Maximisation model, as recommended by De la Barra (1989). The input data for this model was the number of trips that originated and terminated in each zone. The model calculates the elements of the O-D matrix by maximising entropy or 'disorder'. Let 0, be the number of freight transport trips with origin in zone i, and let Dj be the number of freight transport trips with destination in zoney. The Enthropy Maximisation model is then written as follows:
(1) (2) (3) (4)
This is a non-linear optimisation problem with linear constraints. Due to the large number of variables (1292, with 129 being the number of zones in the city of Seville), the resolution technique for this problem was based on Frank-Wolfe's linear approximation (Bertsekas, 1995), which includes the following steps: a) Calculation of a starting point x°: a different problem is solved, where the same constraints are maintained but the objective function is a different one. In particular, the general problem: Min f(x (5) s.t.: at'x = b,, (6) x,A),i = l...n. (7)
72
Logistics systems for sustainable cities
Can be replaced by: Min s.t:
cy a,'x+y = bi, x,A), y/0, i = l...n.
(8) (9) (10)
Where c is a positive constant. The optimal solution for this problem is the starting point x°, b) Solving of the following linear problem: Max
V(w(xo))-x
(11)
s.t:
Same restrictions
(12)
Let xa* be the optimal solution to this problem. c) Uni-directional linear maximisation of: x°+A(x'a-x°)
.with As [0,1]
(13)
d) If the problem does not converge (that is, if :/is not close enough to x°), go back to step to with x° = /• This technique was applied to the estimation of O-D freight matrices for home deliveries, and for each one of the five sectors considered in the analysis.
HOME DELIVERIES ESTIMATION The number of origins and destinations for each zone are calculated as weighted averages: (15)
where: R, is the number of retailers doing home deliveries in zone /, that is, those dealing with furniture (sector 4), home appliances (sector 2), office appliances (sector 2), and doors, windows and blinds (sector 2). Nd is the average number of freight vehicles used daily for home deliveries, equal to 1,028. As stated by the results of the surveys carried out, this is equal to Rt. Pi is the population in zone i. An entropy maximisation model was thus formulated for this data, and the corresponding O-D matrix for home deliveries was estimated.
Estimation of an O-D matrix for freight transport
73
DELIVERIES TO RETAILERS For each sector k and for each destination zone j the number of daily deliveries xkj can be estimated as follows: (16) where: dj is the number of freight-receiving retailers of type / (where / is each one of the types of retailers shown in Table 2), included in sector k, located in zone/. This data is contained in the information supplied by the Chamber of Commerce. e1 is the average number of deliveries received daily by a generic retailer of type /. This data is taken directly from Table 2. The values thus calculated for the **• are shown in Table 3. Once the number of daily deliveries for each sector in each zone of the city has been calculated, it is possible to estimate the number of vehicles dedicated to them as follows:
where, Dkj - number of freight vehicles of sector k which have their destination in zone/ N = number of freight vehicles doing deliveries from wholesalers to retailers in the city. xkj = number of daily deliveries received by sector k in zoney (Table 3). It is important to mention that calculations are performed in a different manner for sector 5, dedicated to fresh food products. These products are distributed daily for the whole city from the Seville Fresh Food Wholesaler, Mercasevilla, which is located in zone 104. The number of vehicles with a license for operating from Mercasevilla is 206, which means that for sector 5 there will be 206 vehicles with origin in zone 104, and their destination zones will be calculated according to equation (17). Then, for the other four sectors, the remaining number of vehicles will be: N = 5,823 -Nd- 206 = 4,589. The values for fl* are shown in Table 4.
74
Logistics systems for sustainable cities Table 3 Average number of deliveries received daily in each zone of the city
o S 1 2 3
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
SECTOR 1 11.4 11.4 11.4 8.9 8 8 11.4 8 20.1 20.1 8.9 8.9 8.9 8.9 602 602 602 602 602 2.3 602 2.3 602 6 6 602 6 6 6 602 602 6 6 6 6 6
6 6 6 8.9 8.9 6
6
2 26.3 26.3 26.3 16.3 28.2 28.2 26.3 28.2 42.7 42.7 16.3 16.3 16.3 16.3 7.8 7.8 7.8 7.8 7.8 3.8 7.8 3.8 7.8 14.2 14.2 7.8 14.2 14.2 14.2 7.8 7.8 14.2 14.2 14.2 14.2 14.2 14.2 14.2 14.2 16.3 16.3 14.2 14.2
3 201.5 201.5 201.5 145.3 174.5 174.5 201.5 174.5 296 296 145.3 145.3 145.3 145.3 87.3 87.3 87.3 87.3 87.3 26 87.3 26 87.3 114.6 114.6 87.3 114.6 114.6 114.6 87.3 87.3 114.6 114.6 114.6 114.6 114.6 114.6 114.6 114.6 145.3 145.3 114.6 114.6
4
5
75.1 75.1 75.1 63 103 103 75.1 103 93.7 93.7 63 63 63 63 27.3 27.3 27.3 27.3 27.3 4.3 27.3 4.3 27.3 38.8 38.8 27.3 38.8 38.8 38.8 27.3 27.3 38.8 38.8 38.8 38.8 38.8 38.8 38.8 38.8 63 63 38.8 38.8
8 8 8 14.6 10.3 10.3 8 10.3 20.5 20.5 14.6 14.6 14.6 14.6 7.6 7.6 7.6 7.6 7.6 3 7.6 3 7.6 12.7 12.7 7.6 12.7 12.7 12.7 7.6 7.6 12.7 12.7 12.7 12.7 12.7 12.7 12.7 12.7 14.6 14.6 12.7 12.7
o SI 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64
1
2
SECTOR 3
14.2 6 6.7 17.8 6.7 17.8 6.7 17.8 6.7 17.8 6.7 17.8 6.7 17.8 6.7 17.8 6.7 17.8 28.2 8 13.5 14.3 13.5 14.3 13.5 14.3 13.5 14.3 13.5 14.3 13.5 14.3 13.5 14.3 13.5 14.3 13.5 14.3 16.2 17.1 11.2 16.9 65 11.2 16.9 66 11.2 16.9 67 11.2 16.9 68 11.2 16.9 69 11.2 16.9 70 11.2 16.9 71 11.2 16.9 72 11.2 16.9 73 11.2 16.9 74 11.2 16.9 4.9 75 5.1 4.9 76 5.1 4.9 77 5.1 4.9 78 5.1 79 5.1 4.9 80 5.1 4.9 81 5.1 4.9 82 13.5 14.3 83 5.1 4.9 84 5.1 4.9 85 13.5 14.3 86 5.1 4.9
114.6 107.8 107.8 107.8 107.8 107.8 107.8 107.8 107.8 174.5 105.8 105.8 105.8 105.8 105.8 105.8 105.8 105.8 105.8 126.9 137.8 137.8 137.8 137.8 137.8 137.8 137.8 137.8 137.8 137.8 137.8 58.2 58.2 58.2 58.2 58.2 58.2 58.2 105.8 58.2 58.2 105.8 58.2
4
5
38.8 33.4 33.4 33.4 33.4 33.4 33.4 33.4 33.4 103 37.8 37.8 37.8 37.8 37.8 37.8 37.8 37.8 37.8 45.4 40.1 40.1 40.1 40.1 40.1 40.1 40.1 40.1 40.1 40.1 40.1 16.5 16.5 16.5 16.5 16.5 16.5 16.5 37.8 16.5 16.5 37.8 16.5
12.7 7.9 7.9 7.9 7.9 7.9 7.9 7.9 7.9 10.3 9.8 9.8 9.8 9.8 9.8 9.8 9.8 9.8 9.8 11.7 18 18 18 18 18 18 18 18 18 18 18 8.2 8.2 8.2 8.2 8.2 8.2 8.2 9.8 8.2 8.2 9.8 8.2
.3 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
122 123 124 125 126 127 128 129
1
2
5.1 5.1 5.1 5.1 6 7.9 7.9 7.9 7.3 7.3 7.3 7.3 7.3 7.3 7.3 7.3 12 12 12 12 12 12 12 24.2 2 2 9.7
4.9 4.9 4.9 4.9 10.9 16.1 16.1 16.1 15 15 15 15 15 15 15 15 22.3 22.3 22.3 22.3 22.3 22.3 22.3 41 13.4 13.4 11.8 11.8 11.8 17.8 11.8 47.2 11.8 11.8 11.8 0.6 10.9 0.6 10.9 10.9 10.9 26 10.9
9.7 9.7 6.7 9.7 19.4 9.7 9.7 9.7 1 6 1 6
6 6 16.2 6
SECTOR 4 3 58.2 58.2 58.2 58.2 44.2 127 127 127 108.6 108.6 108.6 108.6 108.6 108.6 108.6 108.6 191.6 191.6 191.6 191.6 191.6 191.6 191.6 351 52.5 52.5 96 96 96 107.8 96 155 96 96 96 31.5 44.2 31.5 44.2 44.2 44.2
221 44.2
16.5 16.5 16.5 16.5 12.5 31.4 31.4 31.4 42.6 42.6 42.6 42.6 42.6 42.6 42.6 42.6 54 54 54 54 54 54 54 119.4 8.6 8.6 33.2 33.2 33.2 33.4 33.2 30.2 33.2 33.2 33.2 2.3 12.5 2.3 12.5 12.5 12.5 70.6 12.5
5 8.2 8.2 8.2 8.2 4.3 5 5 5 6
6 6
6 6
6 6 6 23 23 23 23 23 23 23 15 0 0 5.1 5.1 5.1 7.9 5.1 28 5.1 5.1 5.1 1.5 4.3 1.5 4.3 4.3 4.3 24 4.3
Estimation of an O-D matrix for freight transport
75
2
1.73 1.73 1.73 1.35 1.22 1.22 1.73 1.22 3.06 3.06 1.35 1.35 1.35 1.35 91.58 91.58 91.58 91.58 91.58 0.35 91.58 0.35 91.58 0.91 0.91 91.58 0.91 0.91 0.91 91.58 91.58 0.91 0.91 0.91 0.91 0.91 0.91 0.91 0.91 1.35 1.35 0.91 0.91
4.00 4.00 4.00 2.48 4.29 4.29 4.00 4.29 6.50 6.50 2.48 2.48 2.48 2.48 1.19 1.19 1.19 1.19 1.19 0.58 1.19 0.58 1.19 2.16 2.16 1.19 2.16 2.16 2.16 1.19 1.19 2.16 2.16 2.16 2.16 2.16 2.16 2.16 2.16 2.48 2.48 2.16 2.16
3 30.65 30.65 30.65 22.10 26.55 26.55 30.65 26.55 45.03 45.03 22.10 22.10 22.10 22.10 13.28 13.28 13.28 13.28 13.28 3.96 13.28 3.96 13.28 17.43 17.43 13.28 17.43 17.43 17.43 13.28 13.28 17.43 17.43 17.43 17.43 17.43 17.43 17.43 17.43 22.10 22.10 17.43 17.43
4 11.43 11.43 11.43 9.58 15.67 15.67 11.43 15.67 14.25 14.25 9.58 9.58 9.58 9.58 4.15 4.15 4.15 4.15 4.15 0.65 4.15 0.65 4.15 5.90 5.90 4.15 5.90 5.90 5.90 4.15 4.15 5.90 5.90 5.90 5.90 5.90 5.90 5.90 5.90 9.58 9.58 5.90 5.90
5 1.22 1.22 1.22 2.22 1.57 1.57 1.22 1.57 3.12 3.12 2.22 2.22 2.22 2.22 1.16 1.16 1.16 1.16 1.16 0.46 1.16 0.46 1.16 1.93 1.93 1.16 1.93 1.93 1.93 1.16 1.16 1.93 1.93 1.93 1.93 1.93 1.93 1.93 1.93 2.22 2.22 1.93 1.93
44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
1 0.91 1.02 1.02 1.02 1.02 1.02 1.02 1.02 1.02 1.22 2.05 2.05 2.05 2.05 2.05 2.05 2.05 2.05 2.05 2.46 1.70 1.70 1.70 1.70 1.70 1.70 1.70 1.70 1.70 1.70 1.70 0.78 0.78 0.78 0.78 0.78 0.78 0.78 2.05 0.78 0.78 2.05 0.78
2
SECTOR 4 3
2.16 17.43 2.71 16.40 2.71 16.40 2.71 16.40 2.71 16.40 2.71 16.40 2.71 16.40 2.71 16.40 2.71 16.40 4.29 26.55 2.18 16.10 2.18 16.10 2.18 16.10 2.18 16.10 2.18 16.10 2.18 16.10 2.18 16.10 2.18 16.10 2.18 16.10 2.60 19.31 2.57 20.96 2.57 20.96 2.57 20.96 2.57 20.96 2.57 20.96 2.57 20.96 2.57 20.96 2.57 20.96 2.57 20.96 2.57 20.96 2.57 20.96 0.75 8.85 0.75 8.85 0.75 8.85 0.75 8.85 0.75 8.85 0.751 8.85 0.75 8.85 2.18 16.10 0.75 8.85 0.75 8.85 2.18 16.10 0.75 8.85
5.90 5.08 5.08 5.08 5.08 5.08 5.08 5.08 5.08 15.67 5.75 5.75 5.75 5.75 5.75 5.75 5.75 5.75 5.75 6.91 6.10 6.10 6.10 6.10 6.10 6.10 6.10 6.10 6.10 6.10 6.10 2.51 2.51 2.51 2.51 2.51 2.51 2.51 5.75 2.51 2.51 5.75 2.51
5 1.93 1.20 1.20 1.20 1.20 1.20 1.20 1.20 1.20 1.57 1.49 1.49 1.49 1.49 1.49 1.49 1.49 1.49 1.49 1.78 2.74 2.74 2.74 2.74 2.74 2.74 2.74 2.74 2.74 2.74 2.74 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.49 1.25 1.25 1.49 1.25
NOZ
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
SECTOR 1
NOZ
ZON
Table 4 Total number of trips with destination on each zone of the city
1 87 0.78 88 0.78 89 0.78 90 0.78 91 0.91 92 1.20 93 1.20 94 1.20 95 1.11 96 1.11 97 1.11 98 1.11 99 1.11 100 1.11 101 1.11 102 1.11 103 1.83 104 1.83 105 1.83 106 1.83 107 1.83 108 1.83 109 1.83 110 3.68 111 0.30 112 0.30 113 1.48 114 1.48 115 1.48 116 1.02 117 1.48 118 2.95 119 1.48 120 1.48 121 1.48 122 0.15 123 0.91 124 0.15 125 0.91 126 0.91 127 0.91 128 2.46 129 0.91
SECTOR 4 3
2 0.75 0.75 0.75 0.75 1.66 2.45 2.45 2.45 2.28 2.28 2.28 2.28 2.28 2.28 2.28 2.28 3.39 3.39 3.39 3.39 3.39 3.39 3.39 6.24 2.04 2.04 1.80 1.80 1.80 2.71 1.80 7.18 1.80 1.80 1.80 0.09 1.66 0.09 1.66 1.66 1.66 3.96 1.66
8.85 8.85 8.85 8.85 6.72 19.32 19.32 19.32 16.52 16.52 16.52 16.52 16.52 16.52 16.52 16.52 29.15 29.15 29.15 29.15 29.15 29.15 29.15 53.40 7.99 7.99 14.60 14.60 14.60 16.40 14.60 23.58 14.60 14.60 14.60 4.79 6.72 4.79 6.72 6.72 6.72 33.62 6.72
2.51 2.51 2.51 2.51 1.90 4.78 4.78 4.78 6.48 6.48 6.48 6.48 6.48 6.48 6.48 6.48 8.22 8.22 8.22 8.22 8.22 8.22 8.22 18.16 1.31 1.31 5.05 5.05 5.05 5.08 5.05 4.59 5.05 5.05 5.05 0.35 1.90 0.35 1.90 1.90 1.90 10.74 1.90
5 1.25 1.25 1.25 1.25 0.65 0.76 0.76 0.76 0.91 0.91 0.91 0.91 0.91 0.91 0.91 0.91 3.50 3.50 3.50 3.50 3.50 3.50 3.50 2.28 0.00 0.00 0.78 0.78 0.78 1.20 0.78 4.26 0.78 0.78 0.78 0.23 0.65 0.23 0.65 0.65 0.65 3.65 0.65
Once the /_)* values are known, it is possible to estimate the number of vehicles of each sector k with origin in each zone i, that is, the O*, for freight transport between wholesalers and retailers. The starting assumption states that the number of trips originated in the city for each sector O* must equal the number of destinations: (18)
76 Logistics systems for sustainable cities This is the total number of vehicles assigned to the wholesalers in each sector for the whole city. The amount of these vehicles that have their origin in each zone of the city is then estimated as follows: (19)
where: O,k - number of freight vehicles of sector k that have their origin in zone i (f = total number of vehicles for sector k of = number of wholesalers of sector k located in zone ;' The calculated of values are shown in Table 5.
Estimation of an O-D matrix for freight transport 77
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
2.75 2.75 2.75 2.90 0.00 0.00 2.75 0.00 3.05 3.05 2.90 2.90 2.90 2.90 3.51 3.66 3.81 3.97 4.12 19.84 4.43 21.36 4.73 5.34 5.34 5.34 5.34 5.34 5.34 5.80 5.95 5.34 5.34 5.34 5.34 5.34 5.34 5.34 5.34 2.90 2.90 5.34 5.34
4
5
9.23 9.23 9.23 7.81 11.54 11.54 9.23 11.54 15.66 15.66 7.81 7.81 7.81 7.81 2.97 3.02 3.06 3.11 3.15 12.87 3.24 13.31 3.33 4.53 4.53 4.53 4.53 4.53 4.53 3.64 3.68 4.53 4.53 4.53 4.53 4.53 4.53 4.53 4.53 7.81 7.81 4.53 4.53
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
SECTOR 2
3
0.15 3.66 0.15 3.66 0.15 3.66 0.12 2.53 0.09 2.82 0.09 2.82 0.15 3.66 0.09 2.82 1.54 19.71 1.54 19.71 0.12 2.53 0.12 2.53 0.12 2.53 0.12 2.53 1.14 12.11 1.17 12.39 1.20 12.67 1.23 12.95 1.26 13.23 5.86 68.99 1.33 13.80 6.16 71.80 1.39 14.36 2.40 13.23 2.40 13.23 2.40 13.23 2.40 13.23 2.40 13.23 2.40 13.23 1.60 16.33 1.63 16.61 2.40 13.23 2.40 13.23 2.40 13.23 2.40 13.23 2.40 13.23 2.40 13.23 2.40 13.23 2.40 13.23 0.12 2.53 0.12 2.53 2.40 13.23 2.40113.23
3
4
5
ZONE
1
ZONE
ZONE
Table 5 Total number of trips with origin on each zone of the city
1
13.23 47.59 47.59 47.59 47.59 47.59 47.59 47.59 47.59 2.82 18.30 18.30 18.30 18.30 18.30 18.30 18.30 18.30 18.30 18.30 20.27 20.27 20.27 20.27 20.27 20.27 20.27 20.27 20.27 20.27 20.27 4.79 4.79 4.79 4.79 4.79 4.79 4.79 18.30 4.79 4.79 18.30 4.79
4.53 8.34 8.34 8.34 8.34 8.34 8.34 8.34 8.34 11.54 5.77 5.77 5.77 5.77 5.77 5.77 5.77 5.77 5.77 5.77 6.48 6.48 6.48 6.48 6.48 6.48 6.48 6.48 6.48 6.48 6.48 1.20 1.20 1.20 1.20 1.20 1.20 1.20 5.77 1.20 1.20 5.77 1.20
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129
3.36 3.36 3.36 3.36 15.72 17.24 17.24 17.24 8.54 8.54 8.54 8.54 8.54 8.54 8.54 8.54 10.22 10.22 10.22 10.22 10.22 10.22 10.22 6.10 3.05 3.05 13.28 13.28 13.28 20.90 13.28 3.05 13.28 13.28 13.28 0.00 15.72 0.00 15.72 15.72 15.72 6.10 15.72
SECTOR 1
2
5.34 2.40 20.90 7.80 20.90 7.80 20.90 7.80 20.90 7.80 20.90 7.80 20.90 7.80 20.90 7.80 20.90 7.80 0.00 0.09 8.01 2.23 8.01 2.23 8.01 2.23 8.01 2.23 8.01 2.23 8.01 2.23 8.01 2.23 8.01 2.23 8.01 2.23 8.01 2.23 12.05 3.51 12.05 3.51 12.05 3.51 12.05 3.51 12.05 3.51 12.05 3.51 12.05 3.51 12.05 3.51 12.05 3.51 12.05 3.51 12.05 3.51 3.36 0.83 3.36 0.83 3.36 !0.83 3.36 0.83 3.36 iO.83 3.36 ;0.83 3.36 0.83 8.01 2.23 3.36 0.83 3.36 0.83 8.01 2.23 3.36 0.83
SECTOR 2
3
4
5
0.83 0.83 0.83 0.83 1.60 2.37 2.37 2.37 3.05 3.05 3.05 3.05 3.05 3.05 3.05 3.05 2.37 2.37 2.37 2.37 2.37 2.37 2.37 4.01 1.23 1.23 1.54 1.54 1.54 7.80 1.54 0.62 1.54 1.54 1.54 0.00 1.60 0.00 1.60 1.60 1.60 1.54 1.60
4.79 4.79 4.79 4.79 27.59 18.87 18.87 18.87 15.49 15.49 15.49 15.49 15.49 15.49 15.49 15.49 16.05 16.05 16.05 16.05 16.05 16.05 16.05 118.2 0.00 0.00 18.02 18.02 18.02 47.59 18.02 19.71 18.02 18.02 18.02 2.82 27.59 2.82 27.59 27.59 27.59 25.34 27.59
1.20 1.20 1.20 1.20 5.90 2.66 2.66 2.66 5.46 5.46 5.46 5.46 5.46 5.46 5.46 5.46 6.52 6.52 6.52 6.52 6.52 6.52 6.52 22.63 1.77 1.77 9.45 9.45 9.45 8.34 9.45 3.11 9.45 9.45 9.45 0.00 5.90 0.00 5.90 5.90 5.90 4.88 5.90
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 206 0 0 0 0 0 0 0 0 0 0 0 0 0
Different entropy maximisation models were formulated for each sector, thus obtaining five new O-D matrices for deliveries to retailers in the city.
78
Logistics systems for sustainable cities
RESULTS AND MODEL VALIDATION Once the six O-D matrices, one for home deliveries and five more for the different retail sectors, were estimated, they were used as input for the EMME/2 commercial simulation package, in order to carry out a private-freight traffic conjoint assignment. The O-D matrix for private mobility in Seville had already been introduced in the package. The results obtained from the simulation were passenger automobile traffic and freight vehicle flows for all the links in the street network of the city. Figure 1 shows details corresponding to the surroundings of the Historical Centre, with both the private cars and estimated freight vehicle flows depicted on it.
Figure 1 Traffic volumes for private cars and freight vehicles obtained by EMME/2 for the surroundings of the historical centre of Seville. The validation of the obtained results was carried out by comparison with actual freight vehicle flows measured at different points of the city. The points that were chosen all correspond to first-order streets or avenues, and are distributed along the city as shown in Figure 2. Table 6 contains the comparison between the estimated and actual freight vehicle flows for the selected locations in Figure 2.
Estimation of an O-D matrix for freight transport
79
Figure 2 Location of freight traffic counts distributed along the city of Seville
80 Logistics systems for sustainable cities Table 6 Comparison between the results of the model and actual vehicle counts in different parts of the city No
Location
Direction
1
Cristo Expiration Bridge
2
Inner Ring (Pasarela)
3
Luis Montoto
4
Colon Avenue
5 6 7 8 9 10 11
Eduardo Dato Palmera Avenue Kansas City San Telmo Bridge Rep. Argentina Avenue Inner Ring (Trinidad) Puerta Carmona
Into the Centre Clockwise Anti-clockwise Into the Centre North South Into the Centre Into the Centre Into the Centre Out of the Centre Out of the Centre Clockwise Into the Centre
Freight vehicles Freight vehicles per hour (Model) per hour (Real) 14 54 23 30 29 17 31 29 7 63 24 34 29
60 60 18 72 66 30 36 84 72 54 36 42 54
The comparison presented in Table 6 shows significant differences between actual freight vehicle flows and those estimated by the model, but these differences can be explained: (a) Values measured from reality tried to include only licensed freight vehicles, but also nonlicensed ones (a significant amount, according to FATRANS), may have been counted. Licensed freight vehicles carry a special sign, but sometimes when a vehicle has been licensed but is not licensed any more, the sign is not removed. (b) Some of the vehicles counted may not correspond to vehicles driving to their destination zone, but to vehicles following their route within the destination zone. (c) At count points 1, 6 and 7, the estimated flow values are extremely lower than 50% of the actual values. But these points are on the main roads providing access to the city from the West, South and North respectively, which means that the measured flows include a significant amount of vehicles entering the city, that is, inter-urban traffic, which was not considered in the analysis. In order to diminish this error, it would be necessary to extend the model to the whole metropolitan region of Seville, since many of the logistic terminals serving the city are located in the surrounding towns. Therefore, the model should be upgraded to allow it to more accurately predict actual data. The introduction of metropolitan-based freight traffic, following the same steps presented here, seems to be possible, requiring only the collection of secondary data for all towns of the Seville metropolitan area. However, it is easy to note that the second refinement required, the introduction of nonlicensed freight traffic and service vehicles, is not a feasible task, because of the unavailability of any data relating to the amount of the number of vehicles involved. Indirect techniques would thus be required, like the adjustment of O-D matrices following boundary conditions (Spiess, 1990), which modify previously calculated O-D matrices so that they match observed traffic flows on selected links.
Estimation of an O-D matrix for freight transport
81
REFERENCES Bertsekas, D. (1995). Nonlinear programming, Athena Scientific. Cohen, H. (1997). Future directions for freight modeling. Proceedings of the Urban Goods and Freight Forecasting Conference 1995. Federal Highway Administration. De la Barra, T. (1989). Integrated land use and transportation modelling- Cambridge University Press. Mufiuzuri, J. (2003). La logistica urbana de mercancias: soluciones, modelado y evaluation. PhD Thesis, University of Seville. Spiess, H. (1990). A gradient approach for the O-D matrix adjustment problem. EMME/2 Support Centre.
This page is intentionally left blank
6
A NEW INTERACTIVE APPROACH ON ROUTE PLANNING W I T H TIGHT DELIVERY TIME WINDOWS A COMPARISON OF THE OBTAINED PLANNING RESULTS TO OTHER APPROACHES
Dipl. Wirtsch. Ing. Oliver Kunze, PTVAG Karlsruhe, Germany
ABSTRACT Tight delivery time windows together with low margins for delivery services pose a challenge to logistics service providers. This topic is the key to success or failure, especially in the distribution of groceries (ordered on the Internet) to private households. Some business models in the branch assume that a significant number of customers will only order if they are informed during order placement precisely when the goods will be delivered. It is therefore crucial for a good long-term customer relationship that these delivery time windows are reliably kept to. A new approach to match these needs was developed by Heid & Kunze. The results of this approach were put into operation in late 2001. This paper describes the new approach and compares the results with other approaches. The analysis provided in this paper also quantifies the increase in vehicle productivity with the slackening of time constraints
INTRODUCTION One aspect of city logistics is the planning and execution of home delivery. This topic is not new, as home delivery is a service offered in quite a few business types, such as furniture home delivery, white goods home delivery or parcel services. As long as the shipment does not require delivery to the hands of the recipient (e.g. the goods can be put into a letter box), this is not such a major problem.
84 Logistics systems for sustainable cities However, once face to face delivery from the driver to the customer is relevant, the aspect of synchronising the arrival of the delivery with the presence of the receiver at the point of delivery becomes more important. For highly valuable goods (white goods or furniture), customers usually accept a delivery time which has been determined by the shipper. But the less convenient the delivery procedure is for the customer, the less attractive the service of home delivery becomes. A business sector which is very dependent on high service standards in this respect is home grocery delivery (i.e. the fulfilment of eBusiness-based grocery sales). The product value is rather low, face to face delivery is usually a necessity due to the refrigeration needs of the goods1, the volatility of the customer in choosing competitive suppliers is high, and the customer not being at home either causes high cost for a second delivery attempt or may even lead to the loss of the customer for further business. Thus, this market is under high pressure to cut transport costs and use geographical synergies as well as possible, (due to the relatively low order volume) and, on the other hand, requires a service quality in terms of delivery time, flexibility and punctuality, which is not yet offered as a general standard. Two "classic" approaches found in home delivery planning are: (a) The master route plan approach (or in short MRA) A master route plan is filled with the actual orders of a day. Customer requests concerning delivery time windows are not considered at all, or are not precise (e.g. first or second half of the day). After the planning, customers receive a notification of when they will be served. (b) The standard route plan approach (or in short SRA) The daily delivery order volume is handed over to a route optimisation system, which generates routes and calculates the estimated times of arrival. Customer requests concerning delivery time windows are considered as constraints during the route planning) Both approaches allow the customers to be notified of the approximate time of arrival (the second considering customer preferences), but cannot easily cope with the conflicting goals of high service requirements of customers with regard to delivery times and the minimisation of transport costs. As a result, a new online route planning approach (ORA) was implemented to match these high service requirements, whereby the customer is linked online via the Internet to a route planning server, and can thus influence the delivery appointment.
NEW ROUTE PLANNING APPROACH ORA The core idea of the new approach is a route planning server, which: (a) can generate delivery time suggestions based on the actual route plan,
1
If not a dedicated refrigerated service station is being served, as e.g. tower 24. [see www.tower24.de]
Route planning with tight delivery time windows
85
(b) can book the delivery in the delivery time window which the customer has chosen online, i.e. the client is connected to the server during the Internet-based purchasing process2, and (c) needs an initial seed route plan (ISRP) as an external input to get started. This approach covers two important aspects: (a) The client can chose from a list of suggested delivery times and receives a binding confirmation3 of the delivery time when he places the order. (b) The transport costs are kept low, as the system suggests time windows while considering geographical synergies for the route. The approach has been proving its benefits in operational use for more than a year now. The next chapters show the positioning of ORA against SRA based on quantitative tests.
QUANTITATIVE ANALYSIS OF APPROACHES As the customer service time window (STW) is the restricting factor for route building, and the vehicle capacity utilisation is not likely to be anywhere near 100%, capacity constraint checks were omitted in this study. The operational results in the day-to-day use of ORA were satisfactory. Thus this paper is focused on quantifying the results of ORA and comparing them to the results obtainable with MRAandSRA. A set of 500 orders was generated as a benchmark scenario. The location of these orders was randomly assigned to reference points in the street network of Zurich, Switzerland (Figure 1).
Figure 1 Geographical spread of delivery orders For clients without Internet access, a call centre provides online access to the route server. 3
The delivery time window span can be adjusted by a parameter. As a result, a delivery time slot as small as half an
hour can be guaranteed.
86 Logistics systems for sustainable cities The service time requests were randomly generated in a time span between 07:00 and 21:59 according to the pattern shown in Figure 2.
Figure 2 Time spread of order delivery time requests This histogram of a probability density was defined ad hoc on the assumption that a delivery peak before lunchtime and another in the evening can be expected. Basics of the SRA Standard route planning systems generate a route plan for a defined number of orders. If such standard route planning systems are used in an eCommerce environment, it is necessary to impose a filter within the online booking procedure to ensure that no more orders are accepted than can be executed by the transport logistic resources. Thus the SRA needs two components: (a) A filter for acceptance of orders (generated by online customer requests) - e.g. a maximum number of orders per day or a max. number of orders per delivery hour, and (b) a standard route planning as an ex post optimisation tool. As a result, the input for the SRA is a given number of orders, and the output of the SRA is the number of vehicles and routes needed to perform the relevant transport assignments. Algorithmic Route Planning Parameters The standard route planning functionality used for SRA in this analysis utilises an heuristic (based on the savings algorithm). Thus, the parameter settings for the savings alpha and beta (see Neumann and Morlock, 1993: 473) have an effect on the planning results as shown in Table 1.
Route planning with tight delivery time windows 87 Table 1 Decrease of cost driving factors with slackening of service time window width for total number of orders (defined by filter): 280 alpha
radius routes STW width* STW [+/- min] [#] [min] * if not limited by depot opening times
beta
12 11 10 9 8 7 6 12 11 10 9 8 7 6 12 11 10 9 8 7 6
-2 -2 _2 -2 -2 -2 -2 -3 -3 -3 -3 -3 -3 -3
vehicles
m
total [km]
mileage total operation time [hh:mm]
360 360 360 360 360 360 360
180 180 180 180 180 180 180
17 17 19 14 17 17 14
14 12 14 13 12 11 11
461 457 463 429 447 443 468
51:16:00 51:18:00 51:40:00 50:58:00 51:37:00 51:31:00 52:35:00
360 360 360 360 360 360 360 360 360 360 360 360 360 360
180 180 180 180 180 180 180
19 17 20 17 18 14 15
15 12 15 13 13 11 11
469 453 485 443 447 434 467
51:36:00 51:31:00 52:17:00 51:29:00 51:33:00 51:13:00 52:41:00
180 180 180 180 180 180 180
19 18 15 20 16 16 15
16 15 12 13 13 12 12
477 473 462 468 449 434 460
51:48:00 51:55:00 51:21:00 51:54:00 51:32:00 51:14:00 52:11:00
It is not possible to determine an overall "best setting" for these parameters, as the results may generally be improved by individually readjusting these parameters for a dedicated scenario. Thus for further analysis of SRA, the parameter settings of alpha = 7 and beta = -2 were defined ad hoc. Logistics Parameters and Constraints The parameters and constraints that were applied during the evaluation of the SRA are presented in Table 2. Table 2 Parameter settings
88 Logistics systems for sustainable cities Max. driving time Max. waiting time / Stop
8h 60 min
Max. waiting time /Route alpha beta sequence opt.
150 min 7 -2 no
Second delimiter for route length (only driving times) In case a client is close to a route, but the route would arrive earlier than the STW opens, the vehicle might wait idle until the STW starts. This maximum waiting time is considered by the route planning algorithm. Maximum of the sum of idle waiting times per route Internal parameter of savings function used (see preceding chapter) Internal parameter of savings function used An ex post sequence optimisation was not used to ameliorate the routes
Results Obtained by SRA Once the parameters of the system have been set, the major elements influencing the results of SRA are: (a) the width of the service time windows, and (b) the number of orders admitted to the system by the pre-processing filter. Their impact is analysed in below. Influence of Service Time Window It is obvious that the transport costs decrease with the growing radius of the service time windows. Figures 3 and 4 illustrate the impact of the radius of the service time windows on the key indicators for the relevant route plans.
Figure 3 Number of routes & vehicles needed as function of time window radius 4 Total number of orders (defined by filter): 140 out of 500
4
Further analysis has shown that the decrease in number of routes and vehicles is not continuous with a growing
STW radius, if a savings-based algorithm is used, e.g. an STW of +/- 90 min will result in 12 routes and 7 vehicles.
Route planning with tight delivery time windows
30
60
120
240
89
600
Figure 4 Number of km and operating hours as a function of time window radius Total number of orders (defined by filter): 140 out of 500 Influence of Filter Pre-process The filter pre-process determines the number of orders accepted by the system. As expected, a decrease in the number of vehicles, the total delivery time and the total delivery route mileage occurs in every scenario (Table 3). Table 3 Decrease in cost driving factors with slackening of service time window width for total numl ler of orde;rs (defi ined by fil ter): 280 out of '500 STW [min]
width* STl V radius routes [+/ - min] /*1
vt chicles total [km]
P '7
mileage total c operation time [hh:m m]
* if not limited by depot opening times
60 120 240 480 1200
30 60 120 240 600
41 28 21 13 6
17 13 145 11 5
817 672 565 396 292
66:28 61:46 57:20 50:39 46:19
Influence of Order Sequence As each order is piped into the savings based standard route planning module at the same time, the specific sequence of the orders does not influence the results within SRA. Route Structure If not hindered by time constraints, savings based approaches usually create a "flower leaf-kind of route structure (Passions, 1987). In case the time constraints are rather tight, the savings tend to produce a "Makita6"-type route structure (Figure 5).
' Note on slight increase in number of vehicles: growing route length has an impact on possibilities for vehicle re-assignment within the given driving time constraints. Thus less (but longer) routes can result in more vehicles needed, if plans are automatically generated. 6
Makita = Game of "jackstraws" or "pick up sticks"
90 Logistics systems for sustainable cities
Figure 5
left: SWT +/-15 min - "Makita" right: no time constraints - "flower leaf 46 routes, 22 vehicles, 911 km; 68:43 h 6 routes, 4 vehicles, 270 km, 44:53 h Route structure changes with slackening of service time window width; total number of orders (defined by filter): 280
This Makita-type is not necessarily a bad result if subsequent stops in a route are close enough to each other in terms of km and driving time. However, dispatchers prefer a geographically more compact route structure, as this means drivers with a specific knowledge of certain city districts can be deployed. This small disadvantage of "Makita-type routes today can be overcome by means of navigation devices within the vehicle which can be pre-loaded with the stop sequence of a route.
Basics of the MRA The master route approach is as follows: routes are predefined by areas (e.g. ZIP-code7 areas). Thus geographically compact routes can be obtained. The key problem of the MRA is to ensure that the accepted number of deliveries with tight STWs can be executed. One possible approach would be that an upper limit is predefined for the number of orders per master route area. This would be equivalent to an SRA approach per route area. As transport capacities (1 vehicle per master route area) are predefined, geographic imbalances in the demand might result in the rejection of orders for one master route area while a neighbouring route would still contain some unused potential to serve customers. Thus this variant of the MRA approach for eCommerce deliveries with tight time constraints does not seem to be very promising and was not investigated in detail. A second possible approach is to have the same filter as the SRA and build routes in a second phase based on the master route areas (e.g. adjacent ZIP-code areas) instead of free planning e.g. by using the savings algorithm. A sequence optimisation which considers time constraints (i.e. a travelling salesman algorithm which considers time constraints) can help to generate the relevant routes. In the case that geographic imbalances in the order set have to be considered when planning the routes, manual re-dispatching could help to improve the workload balancing of the individual vehicles.
7
ZIP-code = postal code
Route planning with tight delivery time windows
91
Results Obtained by MRA Even if the second MRA approach (see above) seems to be somewhat unsophisticated from an academic point of view, it may be used in practice for eCommerce deliveries with tight time constraints. As the quality of results is very much dependent on the day-to-day order volume and even more on the experience of the dispatcher, a quantitative analysis of the MRA approach would need a long-term statistical study to generate reliable results. As a result, further research on this specific topic still needs to be carried out.
Basics of the ORA As in the case of the SRA, the ORA is also a two-component approach. However the components needed for ORA are not the same as for SRA (see above). The core concept of ORA is to have "an idea of what the final routes should look like". This "idea" is represented by the initial seed route plan (ISRP). A component is therefore needed to generate the ISRP. This component can be a standard route planning system which operates on a specific set of virtual orders. These virtual orders represent the presence of a delivery vehicle in a dedicated region for a dedicated time span (e.g. 90 minutes in ZIP-code area 34560). Based on these virtual orders, the ISRP can be generated using a standard route planning system to generate routes which serve these virtual orders. Such a route of the ISRP might sequentially serve e.g. three to four ZIP-code areas. The ISRP is then used as the basis for the ORA core-component, the ORA-Server, which assigns incoming orders based on the ISRP, and the previously confirmed preceding orders. As a result, the input for the ORA is the ISRP (incl. a predefined number of vehicles), and the output of the ORA is the defined subset of order requests which are accepted by the system together with a dedicated route plan. Results obtained by ORA Once the parameters of the system have been set, the major elements influencing the results of ORA are the initial seed route plan (ISRP), which defines the number of vehicles needed to perform the transport assignments, and the width of the service time windows- Therefore, their impact is analysed below. Influence of ISRP As the ORA is based on an ISRP, the ISRP has a delimiting influence8 on the number of orders accepted by the system. Three ISRPs were compared in this study. These were not specifically adapted to the order demand structure, but were generated on the rather simple assumption that every vehicle should cover approximately three routes per day. An ISRP for 5, 8 and 12 vehicles was created. The ISRP for 8 vehicles and the final bookings based on this ISRP are shown in Figure 6.
8 This delimiter for the SRA is the upper bound of orders per hour defined by the filter.
92 Logistics systems for sustainable cities
The results of the ORA in relation to the relevant ISRPs are shown in Table 4. Table 4 Results of ORA with different ISRPs ; STW: radius [+/-min]; o/v: orders per vehicle STW orders v 5orders o/v ORA v5 orders v ehicles o/v ORA v8 orders v ehicles o/v ORA v12 12 10,42 125 8,50 8 68 13,60 5 68 30 15,08 181 14,63 8 117 19,20 5 96 60 21,00 252 20,25 8 162 30,00 5 150 90 26,58 30,25 8 242 41,40 5 207 120 319 33,75 405 12 48,13 8 385 50,40 5 252 240
S
It is interesting that the best ratio of orders per vehicle is reached with only 5 vehicles. The reason for this seems to be that the routes based on the ISPR with more vehicles are not yet "saturated". Further analysis with more than 500 overall orders will be needed to validate this hypothesis. Influence of Service Time Window As the chart above also shows, the key indicator "orders per vehicle" also generally grows with the STW radius. Influence of Order Sequence The impact of the order sequence on the ORA was also analysed. Only marginally different results were obtained by sorting the orders according to their preferred delivery time (Figure 7).
' In this case every vehicle covered 3 routes.
Route planning with tight delivery time windows
93
Figure 7 Influence of order sequence on the number of orders accepted by ORA sorted: sorted by preferred delivery time / unsorted: randomly piped into ORA
ASSESSMENT OF ORA IN COMPARISON WITH SRA ORA has shown its benefits when used in practice (Table 5). Figure 8 provides a rough comparison of ORA with SRA, based on the key indicator of vehicle productivity "orders/vehicle" (o/v) in relation to the STW.
Figure 8 Comparison of ORA and SRA based on orders /vehicle-indicator - logarithmical10 trends assumed 10
Note: The x & y axis of this chart are inverted.
94 Logistics systems for sustainable cities Table 5 Comparing results of ORA with SRA; STW: radius [+/-min]; o/v: orders per vehicle orders vehicles o/v ORA vl2 orders vehicles o/v OR,4 v8 STW orders Vehicles o/v ORA v5 10,42 12 8,50 125 5 13,60 68 8 30 68 12 15,08 14,63 181 5 19,20 117 8 60 96 252 12 5 21,00 162 20,25 30,00 8 90 150 12 5 26,58 242 30,25 319 41,40 8 120 207 12 33,75 48,13 405 252 50,40 385 8 240 5 STW orders Vehicles o/v SRA o140 orders vehicles o/v SRA o210 orders vehicles o/v SRA o280 14 101 16,47 280 17 12,73 210 15 30 140 21,54 12 17,5 280 13 17,50 210 60 140 8 21 14 20,00 280 23,33 210 10 90 140 6 21 14 20,00 280 20,00 0140 10 120 1420 7 11 25,45 26,25 280 5 28,00 210 8 240 140
This comparison needs to be interpreted carefully due to the following facts. Although both charts refer to service time intervals (STW), the semantics of these STW is different in the context of ORA and the SRA: (a) ORA always gives a precise delivery time at order placement which is not revised. The ORA-STW therefore represents the tolerance of the client to accept a delivery in relation to a preferred point of service time, i.e. a customer would place an order if ORA suggests any delivery within this ORA-STW. However, on placement of the order, the customer will be given a precise delivery time commitment (e.g. +/- 30 min.) no matter how wide the ORA-STW is. (b) SRA works differently. Around the preferred point of service time, the SRA-STW-radius is defined as a possible delivery window. Here, the customer will only be given the SRA-STW as delivery time commitment. Thus, for the same STW stated in the chart below, ORA always commits to a more precise delivery time at order entry, whereas SRA only commits to the STW itself. The implementations of ORA and SRA used for this analysis also use a slightly different approach for distance calculation. For performance reasons, the implementation of ORA uses an approximate distance calculation based on a predefined reference-distance matrix, whereas SRA calculates the exact mileage. The impact of the different distance calculation methods for ORA and SRA has to be quantified in detail, in order to obtain more accurate results when comparing ORA with SRA. The key results from Figure 8 are: (a) Both approaches show a strong sensitivity of the cost indicator orders/vehicle to STW-radius (b) The SRA-o/v figures show a lower sensitivity to delivery STW-radius than the ORA-o/v figures. Another important aspect in the home delivery business is the restricted fleet size". Therefore, if more orders are accepted than can be executed with the available fleet, a significant problem of fulfilment occurs. The ORA will ensure that any given number of vehicles will not be exceeded for the execution of the amount of orders accepted by the system. With SRA there is a risk that there is no chance of achieving the accepted number of orders with a given number of vehicles. 11
For food home delivery services usually special small vans with zone refrigeration are used.
Route planning with tight delivery time windows
95
As the SRA-filter can only constrain the number of orders, but does not know how many vehicles finally are needed to execute the accepted orders, the SRA-filter must be carefully adjusted. This fact has to be considered when interpreting the above chart. This filter has a significant impact on the productivity, particularly in the case of a wide STW-radius. If the filter had been less restrictive and had allowed more orders to pass through, it may still have been possible to execute these orders with the same amount of vehicles. Assuming that the goal of the company which runs such a home delivery service is to operate with a fairly small number of vehicles, one advantage of the ORA becomes evident. For example, the goal is to execute 140 orders with 5 vehicles (i.e. an average of 28 orders per vehicle). The ORA will ensure that not more than the 5 available vehicles will actually be needed. The SRA will ensure that a maximum of 140 orders will be accepted. For this case, simulation shows that ORA could grant a service time reliability of +/- 1.5 hours, whereas SRA could only grant a service time reliability of +/- 4 hours12 (Figure 9).
Figure 9 Cost-Service analysis for a small number of available vehicles (excerpt of chart 8) Finally, a non-quantitative advantage of ORA compared to SRA should be pointed out. The ORA will ensure, that drivers operate in a predefined area. Productivity losses due to a lack of driver orientation can therefore be avoided.
OUTLOOK The ORA is dependent on: (a) search strategies for the optimum insertion of new stops into existing routes and their parameters, (b) the initial seed route plans (ISRP), and As discussed above, relaxing the SRA-filter might have improved the SRA-o/v-figures in this example on the cost of running a higher fulfilment risk.
96
Logistics systems for sustainable cities
(c) the accuracy of the anticipation of the number of orders which are to be dealt with (example: if the initial seed route plan is designed for 100 deliveries, but 1000 orders are placed, the result will be considerably different from a result given by a seed route plan for 1000 deliveries), The overall parameterisation of the ORA and its effects on costs and the obtainable service levels need to be further analysed in depth. Further research will also analyse the impact on variable costs (especially driving hours and km) of the different approaches.
ACKNOWLEDGEMENTS The author would like to thank PTV AG for the support provided by agreeing to allow its tools to be used for this research. Special thanks go to my colleagues Werner Heid, who designed and implemented the ORA-functionality, Frank Radaschewski, Marcus Eiser and Mario Holder, who supported me on all set-up and parameterisation topics and to Hansjorg Back for his valuable feedback.
REFERENCES Passions, H. (1987). Tourenplanung mit Tour Master, Oldenbourg, Munchen. Neumann, M. (1993). Operations Research, Carl Hanser, Munchen.
APPENDIX Glossary & Abbreviations Abbreviation ISRP MRA ORA SRA STW o/v
Explanation initial seed route plan (needed as basis for ORA) master route planning approach - assigning stops to routes based on master routes only online route planning approach - assigning stops to routes online based on the ISRP and the preceding confirmed orders standard route plan approach - assigning stops to routes using standard route planning functionality (Savings) service time windows - timeslot for a delivery order, within which a delivery can take place, e.g. 08:00-08:30 Orders per vehicle —indicator of the productivity of vehicles used for eCommerce deliveries
7
INTELLIGENT VEHICLE ROUTING AND SCHEDULING
Russell G. Thompson, The University of Melbourne, Australia
ABSTRACT Distribution systems are dynamic providing many challenges for traditional optimisation methods. Vehicle routing and scheduling procedures need to be able to cope with uncertainty in order to provide effective decision support systems for fleet managers. This paper presents a new method of representing the uncertainty of travel times between customers based on spare time, that is the amount of time that vehicle arrives at a customer before the end of the time windows. Spare time is incorporated within the optimisation procedures as an indirect benefit. An example was used to estimate the benefits of using the spare-time model compared with the traditional model. Travel speeds were simulated representing a wide range of traffic conditions. Both models had similar performance of both models with good travel time conditions. However, the Spare-Time model had substantially less delay costs incurred in congested traffic conditions. The Spare-Time model was shown to be more robust with respect to travel time conditions than the conventional model. Significant overall savings in direct costs were predicted. Increased levels of service for customers were also achieved.
INTRODUCTION Current vehicle routing and scheduling (VRS) procedures often fail to provide adequate information for distribution managers to assist in lowering their operations costs in dynamic operating environments. Intelligent VRS procedures provide additional information that can assist distribution managers cope with change. The paper will describe the development of a range of procedures designed to provide more intelligence in urban VRS.
98 Logistics systems for sustainable cities Recent developments in artificial intelligence search based techniques such as tabu search allow a much wider range of vehicle routing and scheduling problems to be investigated than traditional optimisation procedures. Typically, there are a number of assumptions in current VRS procedures, such as: (a) fixed time windows (non-negotiable) (b) fixed fleets (cannot substitute other vehicles with different capacity, operating costs, fuel consumption, emissions,...) (c) all customers must be serviced (no subcontracting) (d) only one vehicle can service a customers request (load splitting is not permitted) (e) customers must be served directly (no intermediate storage locations can be used) (f) travel times between customers are certain (not stochastic) (g) all orders must be received before vehicles are dispatched (no new jobs allowed during the run) (h) only financial objectives (not environmental or level of service objectives) While the above assumptions may be reasonable in many situations they are often not realistic or practical in contemporary logistics chains. Due to the dynamics of urban distribution systems there is a need to develop more flexible vehicle routing and scheduling procedures to minimise transport costs. Intelligence involves being: (a) Clever - successful in coping with new situations and solving problems (b) Alert - fast in perceiving and understanding problems and in finding answers to challenges The next generation of vehicle routing and scheduling procedures needs to provide more information for managers to enable them to cope with new situations. More flexible and responsive decision support systems should be developed. Variable nature of travel times Existing procedures do not account for the uncertainty and variability of travel times experienced by vehicles operating in urban traffic networks. Consequently, customers can often experience delays when receiving goods due to unexpected congestion being encountered by carriers. A new procedure based on the concept of spare time - the amount of time a vehicle is predicted to arrive at a customer before the end of the time window is used to determine more robust solutions that are better able to cope with unexpected traffic congestion than current methods. Soft time windows allow vehicles to arrive at customers outside of customer specified time windows. Penalties are generally assumed to be reduce the occurrence of such situations. It is common to assume that vehicles have to wait at a customer before being serviced if they arrive before the start of the time window. Vehicle routing and scheduling formulations typically include travel time costs as well as penalty costs associated with vehicles not arriving at customers during time windows. However, travel times are considered deterministically and often there is considerable different between the predicted travel times between customers and those realised. Conventional procedures have
Intelligent vehicle routing and scheduling
99
difficulty incorporating such travel time variations and thus the performance of urban distribution operations maybe quite different than those predicted. A new approach has been developed in an attempt to overcome and improve the performance in light of the uncertainties associated with travel times. Here, spare time is defined as the length of time a vehicle arrives at a customer before the end of its time window. It has been incorporated within the objective function for the vehicle routing problem with soft time windows Contemporary distribution systems frequently involve carriers having to deliver goods within narrow time windows specified by customers. Current vehicle routing and scheduling procedures typically do not account for the variability of travel times. This paper presents a model that takes into account the uncertainty of travel time between customers by incorporating the spare time - the amount of time vehicles are predicted to arrive at a customer before the end of the specified time windows. The benefits of using this approach is illustrated using a small problem based on the distribution of medical products in Melbourne. Significant reductions in costs were estimated. The vehicle routing problem with time windows (VRPTW) can be defined as follows. Let G = Wo, E) be a graph where Uo~ {«», uh -, uN+i) is the vertex set, U= Uo\ {««}, and E = {(«,-, uj): Ui, Uj e Uo, tej] is the arc set. Every vertex of Ucorresponds to a customer to be serviced and uo denotes a depot where vehicles are based. With E is associated a travel time matrix T = (tuj) representing the time taken to travel between vertices for truck X. Vehicle routing and scheduling with soft time windows (VRPSTW) involves penalties being incurred by freight carriers for deliveries outside of customer specified time windows (Taillard et al, 1997). Here, vehicles incur a penalty proportional to the period of time they arrive outside of customers time windows (Figure 1). The VRPSTW consists of finding a permutation n of the sets {X, 0, I,..., N* 0}: X = l,..,m corresponding to tours which start and end at the depot, i.e. n(A, 0) = 0 and n(X, Nn +1) = 0, such that every vertex of V is visited exactly once, the capacity of the trucks is not exceeded and the total routing cost is minimised. where, m - number of trucks available Nx - number of customers visited by truck A (A = l,...,m) For each customer u; the following parameters are specified: (i)
a time window [e/; /,], where e,- and /,- are the earliest and the latest arrival time, respectively, within which the customer should be visited; (ii) a service time xn for loading/unloading the goods from truck X\ (iii) a waiting time unit penalty rate a; for arrival before e,-; (iv) a delay time unit penalty rate fi for arrival after /,-. Although the actual travel time between customers is uncertain in static vehicle routing and scheduling problems a single value estimate (forecast) is usually made (Psaraftis, 1995).
100 Logistics systems for sustainable cities Traditional model The arrival time of a truck at a customer depends on the departure time from the previous customer as well as the travel time from that customer (Equations 1, 2 and 3). a
AMA,i)
=
^A,7t(A,i-V) + tA,n(A,i-V), n(A,i)
0)
where a
A 7t(A i) ~ a r r ' v a l t i m e °f truck X at customer n(X, i)
d^ a^ ,_i) — departure time of truck X from customer n{X, i -1) t& n(A i-l) n(A i) ~ travel time of truck X between customer n{X,i — 1)
and n(X,i) ^AMA.i)
=a
A,n(A,i) + wA,riA,i) + xAMA,i) + bx,x{A,i)
(2)
where
x
A,n(A,i) ~ service (loading/unloading) time of truck X at customer n(X, i),(x^^iAi)
> ^)
^A,7r( A,i) ~ break time of truck X at customer n{X, i), including rest and meal periods
(hMAJ)^0) Most vehicle routing and scheduling models predict the arrival time of trucks at customers deterministically (Equation 3). i-l
i a
AMA,i) = L+2_itAMAJ-l),zU,j)
+
zSw
X,icU,j) +xA,ir(A,j) + ^XMX.j)^
^
where L. - time vehicle X left the depot A
The location and activities associated with the distribution of goods for a single truck can be illustrated using a trajectory diagram (Figure 1).
Intelligent vehicle routing and scheduling
101
Figure 1 A trajectory diagram of a vehicles route With the VRPSTW a penalty cost (PC) incurred for early or late arrival at customers is incorporated within the objective function (Equations 4 and 5). (4)
where (l
if vehicle is used
[0
otherwise
cfx— fixed cost of truck X ($) ek -operating cost for truck X ($/minute)
(5)
102 Logistics systems for sustainable cities Spare-time model The spare time at each customer has been incorporated within the objective function (6). Here, the spare time is considered a benefit.
(6)
(7) where, sx ^
0
— spare time of truck X at customer n(X, i)
Y-nQ-i) spare time ($/hr)
un
it
rate
f ° r arriving at customer 7r(X,i) before the end of its time window
The costs and benefits associated with waiting, spare-time and delay penalty for a vehicle arriving at a customer with a specified time window is illustrated in Figure 2.
Intelligent vehicle routing and scheduling
103
Figure 2 Arrival Related Costs at Customers Benefit analysis A two stage procedure was developed for estimating the benefits (cost savings) of using the model incorporating spare time defined above. The Is' stage involves determining the optimal routes and schedules for both the Traditional and Spare-Time models. The second stage of the process involves testing the performance of the optimal routes and schedules produced by both the Traditional and Spare-Time models under various predicted travel time conditions. This involves generating travel times and evaluating the performance of the optimal routes and schedules generated at the 1st stage using simulation. A costs model is then used to estimate the benefits by comparing the predicted costs of using both types of optimisation models.
Case study A hypothetical distribution problem is presented here to illustrate the differences in performance between the Traditional and Spare-Time models. The problem involves delivering medical supplies from a warehouse to 20 major hospitals in metropolitan Melbourne (Figure 3). Each hospital requires 100kg of medical products to be
104 Logistics systems for sustainable cities delivered within specific time windows (Table 1). Two vans each with a storage capacity of 1000kg are available.
Figure 3 Location of major hospitals in Melbourne Table 1 Customer Time Windows Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Name Monash Medical Centre Sandringham Cabrini Waverley Epworth Mercy Royal Womens Royal Childrens Sunshine Western Frankston NorthPark South Eastern Alfred Box Hill Austin Diamond Valley Lilydale Maroondah Dandenong
After 08:00 08:00 08:00 08:00 08:00 15:00 13:00 10:00 10:00 10:00 10:00 10:00 13:00 13:00 13:00 13:00 15:00 15:00 15:00 15:00
Before 10:00 10:00 10:00 10:00 10:00 17:00 15:00 12 :00 12:00 12:00 12:00 12:00 15:00 15:00 15:00 15:00 17:00 17:00 17:00 17:00
Intelligent vehicle routing and scheduling
105
The values of the cost items for this problem are given in Table 2. Table 2 Costs Parameters (A$/hour) 50 Truck running cost (g^) 50 Waiting time penalty rate (cii) 250 Delay time penalty rate (Pi) 25 Spare time rate (y{) The optimal routes and schedules were determined for both the Traditional and Spare-Time models using the tabu search meta-heuristic technique (Taillard et al, 1997). Based on surveys, a number of parameters were used to estimate the travel time between customers (Table 3). A diversion factor was used to convert the direct distance between customers to travel distances. Table 3 Model parameters Diversion Factor 1.25 Travel Speed (off peak) 45km/h Travel Speed (peak) 30km/h Both models produced optimal solutions with similar direct costs (Table 4) using the parameters in Table 3. The Spare-Time models solution has marginally larger total direct costs but significantly lower travel times as well as substantially larger total waiting and spare time. There were no delay costs for the optimal solutions of either model. Table 4 Comparison of Optimal Total Direct Cost Travel Time (his) ($) Traditional 790.42 11.13 Spare Time 798.78 9.10 Difference (%) 1.06 -18.24
Solutions Waiting Time (hrs) 1.35 3.54 162.22
Spare Time (hrs) 26.93 34.63 28.59
Analysis of the trajectory diagrams of the solutions produced from each model reveals that vans tend to arrive more towards the start of the designated time windows with the optimal solution of the Spare-Time model compared with the Traditional model (Figures 4, 5, 6 and 7). Thus increased travel times between customers would lead to larger delay penalties being incurred when the optimal route of the Traditional model is used compared with Spare-Time model.
106 Logistics systems for sustainable cities
Figure 4 Optimal route from the Traditional model for Van #1
Figure 5 Optimal route from the Spare-Time model for Van #1
Intelligent vehicle routing and scheduling
Figure 6 Optimal route from the Traditional model for Van #2
Figure 7 Optimal route from the Spare-Time model for Van #2
107
108 Logistics systems for sustainable cities Using the process defined above the performance of both optimal solutions was tested when changes in travel speeds were simulated (Figures 8 and 9). The effects on transport costs were estimated when the mean travel speeds were varied. The performance of both models is similar when mean travel speeds were increased. However, considerably less delay time penalties are experienced for the Spare-Time model when the mean travel speeds were reduced (Figures 8 and 9). It can be seen that the Spare-Time model achieves significant savings in costs when the decrease in mean travel speed is more than 20 percent (Figures 10 and 11). The highest savings in percentage terms are achieved when mean travel speeds are reduced by 20 and 30 percent. Lower savings in percentage terms are achieved with lower mean travel speeds since at these levels both models generate large penalty costs. Thus the Spare-Time model performs better in conditions that were moderately more congested than those used to calibrate the models. The expected direct cost savings of using the Spare-Time model compared with the Traditional model is dependent on the travel conditions actually experienced over an extended period. This involves determining the cost savings of using the Spare-Time model for various travel time conditions as well as identifying the likelihood of such conditions being experienced. This can be estimated by multiplying the cost reductions gained by using the Spare-Time model under certain travel time conditions by the probability that those conditions will be realised.
Figure 8 Performance of the Traditional Model
Intelligent vehicle routing and scheduling
Figure 9 Performance of the Spare-Time model
Figure 10 Cost savings for changed travel conditions
Figure 11 Percentage Cost Savings
109
110 Logistics systems for sustainable cities Table 5 presents the estimated savings of using the optimal solution from the Spare-Time model when the mean travel speeds were assumed to be normally distributed. Significant cost reductions were estimated. It can be seen that larger savings would be expected to be realised with a larger variation in the mean travel speeds. Table 5 Expected Benefits Standard Deviation of Saving in total direct costs Mean travel speed (%) (%) 2.3 10 11.9 20 15.6 30
CONCLUSIONS An example was used to illustrate the savings in direct costs that could be expected to be achieved when using the Spare-Time model for delivering goods in urban areas. Larger benefits are realised in conditions that are moderately more congested than those used to calibrate the models. Significant overall savings were estimated from using the Spare-Time model. Increased levels of service for receivers of goods would also be experienced due to the Spare-Time model being more robust in handling changes in travel conditions.
REFERENCES Birge, J.R. and F. Louveaux, (1997). Introduction to Stochastic Programming, Springer, New York. Johnson, N.L. and N.B. Koltz, (1994). Continuous Univariate Distributions, Volume 1, Wiley, New York. Laporte, G., F.V. Louveaux and H. Mercure, (1992). The vehicle routing problem with stochastic travel times, Transportation Science 26, 3, 161-70. Psaraftis, H.N. (1995). Dynamic vehicle routing: Status and prospects, Annals of Operations Research 61, 143-64. Taillard, E., P. Badeau, M. Gendreau, F. Guertin and J.-Y. Potvin, (1997). A tabu search heuristic for the vehicle routing problem with soft time windows, Transportation Science, 31,2,170-186.
8
ROAD NETWORK RELIABILITY ANALYSIS USING VEHICLE ROUTING AND SCHEDULING PROCEDURES
Tadashi Yamada, Hiroshima University, Japan Yohei Yoshimura, Hiroshima University, Japan Kazuhiro Mori, Hiroshima Institute of Technology, Japan
ABSTRACT This paper proposes "Cost Reliability" as a new indicator being useful for assessing the service level of road network. A computer based procedure was developed for estimating the cost distribution and reliability of road network using the vehicle routing and scheduling problem with time windows, and then applied to a test road network with its link travel times being assumed to be normally distributed. Results showed that total cost of each company is not normally distributed due to designated time windows at customers and becomes larger as delay time unit penalty rate increases. This implies that the results from the cost reliability analysis might be different from those from conventional travel time reliability analysis. It is also found out through the road network reliability analyses that probabilistic vehicle routing and scheduling procedures are more robust and performs better than deterministic procedures.
INTRODUCTION There has been a considerable amount of research that has investigated the reliability of transport networks. In systems engineering, reliability may be defined as the degree of stability of the quality of service which a system normally offers. System reliability is becoming increasingly important in the planning, construction and operation of transport networks, if
112 Logistics systems for sustainable cities user demands for high levels of service increase (Bell and Iida, 1997). In general, transport network reliability can be classified into two types, connectivity reliability and travel time reliability (e.g. Asakura and Kashiwadani, 1991; Wakabayashi and Eda, 1992). Connectivity reliability is the probability that traffic can reach a given destination, while travel time reliability is the probability that traffic can reach a given destination within a given time. When there is significant fluctuation in traffic flow, travel times are often longer than expected. As levels of congestion in transport network grow generally, the stability of travel times will have greater significance to transport users. If the probability that drivers can reach their destination from their origin within a desirable time is high, transport network can be considered reliable, in other words, the transport network can provide higher levels of service. Recent advances in vehicle monitoring systems allow information on the variability of travel times to be obtained easily. However, for urban pickup/delivery vehicles, travel time reliability may be insufficient for assessing the service capability of transport network, since vehicles usually make multiple trips and visit many customers. In addition, transport-related companies (i.e. shippers and/or freight carriers) often evaluate the level of service of the transport network not by travel time but by total cost. The total cost incurred for goods pickup/delivery involves the fixed cost of vehicles, vehicle operating cost and early and delay penalty cost at customers. Hence, the fluctuation of these costs plays a vital role in planning routes and schedules of urban pickup/delivery vehicles. This concept of stability for a transport network can be called "Cost Reliability". Cost reliability is similar to travel time reliability in cases where pickup/delivery vehicles have a single OD and customers have no time windows. However, pickup/delivery vehicles often need to visit a number of customers at different locations, and customers often require that they should arrive within specified time windows. In this case, the cost reliability is different from the travel time reliability. Most urban freight transport is undertaken using road-based vehicles. Therefore, this study highlights the cost reliability of road network. Behaviour of pickup/delivery vehicles within urban areas can generally be represented by the vehicle routing and scheduling problem with time windows (VRPTW (e.g. Solomon, 1987)). The VRPTW determines vehicles departure and arrival times at depot and customers as well as the visiting order of customers. These are often determined using a single forecasted travel time of each link on the road network. This implies that link travel times are deterministic. Link travel times are however inherently uncertain, and stochastic variation can be represented based on past experience. The variability of travel times has typically been represented by simple probability distributions. For simplicity, link travel times are assumed to be normally distributed in this paper. A number of activities are involved in urban pickup/delivery of goods, including loading and unloading of goods, travelling between customers and administration at customers. The time taken for these activities is also stochastic in nature. However, this study only considers the uncertainty of travel times between customers. The duration of other activities are assumed to be deterministic. Departure and arrival times at customers are therefore influenced by the stochastic nature of link travel times.
Road network reliability analysis
113
COST DISTRIBUTION PROCEDURE The total cost associated with urban pickup/delivery depends on the departure and arrival times at customers as well as the link travel times. The departure and arrival times at customers also varies due to the stochastic nature of link travel times. The arrival time of a vehicle at a customer is influenced by the departure time at the previous customer. Therefore, cost distributions, which are required for cost reliability analysis, rely on the possible departure and arrival times of all customers. This leads to incorporating the possible departure and arrival times of all customers when estimating cost distributions. However, it is almost impossible to estimate the cost distributions considering such enormous number of combinations within reasonable computation time, and hence this study approximately determines the cost distributions using Monte Carlo simulation. A two-stage procedure was developed for estimating the cost distributions using the VRPTW. The first stage involves determining routes and schedules of vehicles for each freight carrier (Figure 1). Optimal routes and schedules are determined using the Forecasted VRPTW model (Taniguchi et al, 2001). The Forecasted VRPTW model only relies on a single estimate of link travel time. Average link travel time have generally been used in this type of vehicle routing and scheduling procedure. The objective function of the Forecasted VRPTW is represented by the sum of the fixed cost of vehicles, vehicle operating cost and early and delay penalty costs. The fixed cost of vehicles are composed of the vehicle capital cost, tax and insurance costs, ...etc. Vehicle operating costs include the fuel, maintenance, personnel and tolls. Early arrival penalty costs are incurred when arriving at customers earlier than the designated time, whilst the delay time penalty cost is incurred when vehicles arrive late at customers. The Forecasted VRPTW is typically solved under several constraints associated with the capacity of vehicles, the number of visit to each customer, the operating hours for freight carriers and the trips of vehicles (i.e. vehicles have to depart from the depot and return to the depot after visiting customers and are allowed to make multiple routes). Individual freight carriers predict their total costs and operate vehicles on the basis of the optimal routes and schedules being estimated by the Forecasted VRPTW procedure. However, predicted total costs are not always consistent with the total costs actually incurred due to the variable nature of link travel times. The second stage estimates the variability of total cost. Cost distributions can be obtained through this process (Figure 2). Travel times between customers are randomly generated from the probability distribution of each link. These travel times are assumed to be the travel times experienced. An expected total cost is then calculated with the objective function mentioned above. Here, routes and schedules used are identical to the optimal routes and schedules being determined in the first stage. Such calculations for estimating expected total costs are repeatedly performed a total of 100,000 times to determine the cost distributions.
114
Logistics systems for sustainable cities
Figure 1 Route and schedule determination process
Figure 2 Cost distribution estimation process
COST RELIABILITY PROCEDURE The second stage in the cost distribution procedure provides 100,000 different expected total costs for each freight carrier. This also allows the distribution of costs incurred by all freight carriers to be estimated. A total cost for urban pickup/delivery generated on the road network can be estimated with the sum of total cost of individual freight carriers. This also varies within 100,000 times calculation. This variability leads to estimating the distribution of cost generated on the road network. Thus, the cost distribution of road network can be identified. Let Pjj{T) be a probability that a trip between node i andj will be completed within time T, the travel time reliability on the road network is generally defined as follows (Asakura and Kashiwadani, 1991): (1)
Road network reliability analysis
115
where, : travel time between node i andy Therefore, the cost reliability of the road network can also be represented using a similar equation (Equation (2)). (2)
where, (3)
TC
: total cost generated on the road network (= sum of total costs of each freight carrier) TC * : desired value of total cost; given C (t0, X): total cost for freight carrier u
TEST ROAD NETWORK A computer based procedure was developed for estimating the cost reliability of road network and applied to a test road network with uniform distances, where the size of the area is assumed to 400km2 (20 x 20 km) (Figure 3). Link travel times on the road network vary depending on the time period (i.e. peak hours and off-peak hours) and have different means and variances for each time period. The shadowed portion of the road network represents the centre of city where the traffic conditions are relatively congested.
Figure 3 Test road network
116 Logistics systems for sustainable cities Link travel times are assumed to be normally distributed in this study. Standard deviations of link travel times need to be determined as well as mean link travel times. This study adopted the following relationship between mean and standard deviation of travel times (Matsumoto and Shiramizu, 1985). a = 0.360* ft°m
(4)
: standard deviation of link travel time : mean link travel time; given Table 1 Example of customer data (Carrierl &2) Freight carrier Depot node Customer Freight demand Time window number number node number (ton)
1
2
2
3
6 7 8 11 13 13 13 13 16 17 18 7 8 9 13 13 17 18 19 6 7 8 9 13 13 13 13 14 17 18 19 7 8 9 13 13 17 18 19 14
0.1 0.5 0.5 0.1 1 1 1 1 0.1 0.5 0.5 0.5 0.5 0.5 1 1
0.5 0.5 0.5 0.1 0.5 0.5 0.5 1 1 1 1 0.5 0.5 0.5 0.5 0.5 0.5 0.5 1 1 0.5 0.5 0.5 0.5
a.m. no time window no time window 15:00-17:00 9:00-11:00 9:00-11:00 9:00-11:00 9:00-11:00 no time window p.m. 9:00-11:00 no time window no time window p.m. 15:00-17:00 9:00-11:00 9:00-11:00 9:00-11:00 9:00-11:00 no time window a.m. a.m. no time window 9:00-11:00 9:00-11:00 9:00-11:00 9:00-11:00 no time window p.m. 9:00-11:00 15:00-17:00 a.m. a.m. no time window 9:00-11:00 9:00-11:00 no time window no time window no time window p.m.
Road network reliability analysis
117
Eight freight carriers are assumed to be on the road network and operate delivery trucks, each with one depot. Table 1 shows an example of customer data used. Each freight carrier has 20 customers. Customers are mainly distributed within the city centre. Freight demands are larger for customers that are located closer to the centre of city. Designated time windows are categorised into 3 types, time windows of two hours length, time windows for a.m. or p.m. and no time window. Each freight carrier can only operate 2-ton trucks. Unit fixed costs were set at 10418 (yen/vehicle/day). Unit operating costs were set at 14.0 (yen/vehicle/min). Unit penalty rates were distinguished between early and late arrivals. Early arrival time unit penalty rate was assumed to be equivalent to the above unit operating cost, while delay time unit penalty rate was set at 5 times the unit operating cost for 4-ton trucks. These costs are based on results from recent studies of Just-In-Time truck operations in Japan.
COST DISTRIBUTION ANALYSIS Figure 2 displays the cost distributions of freight carrier 2. It can be seen that total cost is not normally distributed (Figure 2(a)). It has a skewed distribution with a longer tail to the right. This is largely due to the penalty costs incurred for late arrivals (Figure 2(c)). Freight carriers tend to avoid incurring such penalty costs. Therefore, delay time penalty costs are not experienced in most cases. Reasons for its longer tail include that the delay time penalty costs are sometimes experienced under the congested conditions where link travel times are relatively large. Penalty costs for late arrivals largely influence the shape of the cost distribution. This is illustrated in Figure 3. Freight carriers experience higher costs, as the delay time unit penalty rate (i.e. cd n(j)) increases. The cost distribution shape is sharper around its mode and its tail also becomes much longer, as the delay time unit penalty rate increases. A test of goodness of fit indicated that the cost distribution without time windows at customers (i.e. in the case of cd n0) =0 in Figure 3) is normal. This implies that the results from the cost reliability analysis might be different from those from conventional travel time reliability analysis under the conditions where time windows are designated by customers.
118 Logistics systems for sustainable cities
Figure 2 Cost distributions (Carrier 2)
Road network reliability analysis
119
Figure 3 Cost distributions comparison (Carrier 2)
RELIABILITY ESTIMATION The cost reliability estimation procedure described above was applied to the test road network. A probabilistic vehicle routing and scheduling procedure was also tested for investigating the influence of using a more advanced vehicle routing and scheduling procedure on road network reliability. The Forecasted VRPTW model determines optimal routes and schedules with a single forecasted value of link travel time. However, the VRPTW models incorporating the stochastic nature of travel times have the potential to reduce transport costs for freight carriers. The Probabilistic VRPTW models have therefore been developed with the uncertainty of predicting travel times being taken into account (Taniguchi et ah, 2001). Travel time distributions between customers are required for computing the expected operating and penalty costs. This distinguishes the Probabilistic VRPTW from the Forecasted VRPTW. The expected vehicle operating cost is calculated as follows in the Probabilistic VRPTW model:
(5)
C,, {t, 0 , x , ) : operating cost for vehicle / (yen) t, 0 x,
: departure time of vehicle / from the depot : assignment and order of visiting customers for vehicle /
c,, : unit operating cost for vehicle / (yen /min); given N, '• total number of customers visited by vehicle / T(tiw,n(i),n(i + l)) : average travel time of vehicle / between customer n{i) and «(; + !)
120 Logistics systems for sustainable cities The Probabilistic VRPTW used here incorporates the variable nature of travel time between customers in estimating expected vehicle arrival times at customers and depot more explicitly than the conventional Probabilistic VRPTW models (Taniguchi et ah, 2001):
(6) (7)
t°nU) : arrival time of vehicle / at customer n(i) tsnU) : start of time window at customer n{i); given p'tn(i)kio-tri(i)'xi) Pin(i)kiO't>xi)
:
Probability that vehicle / leaving the depot at time f ;o will arrive at customer n{i) before time t'n(i) '• probability that vehicle / leaving the depot at time f /o will arrive at customer n(i) at time t
Figure 4 compares the cost distributions of the road network. It can be seen that the cost distribution obtained from using the Probabilistic VRPTW model has a smaller mode, mean and variance. A possible explanation for this is that optimal routes and schedules obtained from the Probabilistic VRPTW procedures tend to avoid late arrivals at customers, since these procedures more explicitly incorporate the risk of late arrivals at customers in determining optimal solutions. Road network reliability analyses can easily be undertaken using the cost distributions of the road network. Given a desired value of total cost, a probability that pickup/delivery activities undertaken on the road network will be completed within the desired value is computed (Equation (2) & Figure 5). Such a probability however depends on the VRPTW procedures used. If the desired value equals 503500 (yen/day), where the two curves cross each other, the difference in the cumulative probabilities estimated from both models becomes largest. (Figure 5). The cost reliability is consistently better when the Probabilistic model is used. This implies that the Probabilistic VRPTW procedure provides better performance as well as that the cost reliability depends on the vehicle routing and scheduling procedures used. Advanced vehicle routing and scheduling procedures could offer the potential to complement the performance of road network without the construction of new roads or increasing road capacity.
Road network reliability analysis
121
Figure 4 Cost distributions of road network
Figure 5 Example of cost reliability estimation
CONCLUSIONS Road network reliability analysis using total costs incurred for urban pickup/delivery would be useful for administrators to assess the service levels of road networks. It has the potential to provide new effective tools for the planning, construction and operation of transport networks. Cost distributions would also be helpful to freight carriers and shippers in finding more stable and cheaper routes and schedules of vehicles, though this depends in part on the vehicle routing and scheduling procedures used. This paper presents procedures for estimating the cost distribution and reliability of road network. These were undertaken on the basis of the variability of link travel times of the road network. Cost distribution analyses for urban goods delivery with a test road network revealed
122 Logistics systems for sustainable cities that total cost of each company is not normally distributed due to designated time windows at customers and becomes larger as delay time unit penalty rate increases. It is also found out through road network reliability analyses that the Probabilistic VRPTW model is more robust and performs better than the Forecasted VRPTW model. Further investigation will however be necessary using the procedures developed with a variety of scenarios relating to new road construction and increased road capacity. The procedures presented here should be applied to actual road networks when investigating the results of such
REFERENCES Asakura, Y. and M. Kashiwadani (1991). Road network reliability caused by daily fluctuation of traffic flow. Proceedings ofl9the PTRC Summer Annual Meeting, 73-84. Bell, M.G.H. and Y. Iida (1997). Transportation Network Analysis, Wiley, West Sussex. Giannopoulos, G.A. and M. McDonald (1997). Developments in transport telematics applications in Japan: traffic management, freight and public transport. Transport Reviews 17(1), 37-59. Matsumoto, S. and Y. Shiramizu (1985). The effect of travel time uncertainty on goods transportation under temporal constraint. Journal of Infrastructure Planning and Management 355 IV-2, 75-82 (in Japanese). Solomon, M. M. (1987). Algorithms for the vehicle routing and scheduling problems with time window constraints. Operations Research, 35, 254-265. Taniguchi, E., Thompson, R.G., Yamada, T. and R. van Duin. (2001). City Logistics: Network Modelling and Intelligent Transport Systems, Pergamon, Oxford. Wakabayashi, H. and Y. Iida (1992). Upper and lower bounds of terminal reliability of road networks: an efficient method with Boolean algebra. Journal of Natural Disaster Science, 14 (1), 29-44. Yamada, T., Taniguchi, E. and Y. Itoh (2001). Co-operative vehicle routing model with optimal location of logistics terminals. In: City Logistics II (E. Taniguchi and R.G. Thompson, eds.), 139-153, Institute for city logistics, Kyoto.
9
O N THE ESTIMATION OF THE MAXIMUM EFFICIENCY OF THE TRUCKING INDUSTRY: IMPLICATIONS FOR CITY LOGISTICS
Jose Holguin-Veras, Ph.D., P.E., Associate Professor, Rensselaer Polytechnic Institute, USA
ABSTRACT This paper presents methodologies for estimating the maximum efficiency that the trucking industry could achieve, given a set of time varying commodity flow matrices. The paper establishes that empty trips are the result of the competition between different segments of the vehicular supply and that the probability of either of these segments transporting a shipment is a function of the percentage of empty trips and the probabilities of pick-ups and deliveries. The research showed that the dynamic relations of supply and demand could be made operational in a simulation system. The quantitative estimates produced here provide an upper bound on the benefits attributable to market efficiency enhancers such as Internet based freight clearinghouses.
INTRODUCTION An area that is receiving increasing attention from researchers is the use of market enhancing mechanisms, such as Internet based freight markets that, by providing a forum where shippers and carriers could (virtually) meet, may enhance the overall efficiency of the trucking system. The fundamental assumption behind these concepts is that increases in market transparency would increase the overall efficiency of the system, to this context, commercial vehicles that otherwise would return empty to their home bases may be able to find suitable cargoes in the virtual freight market. However promising, the systemwide benefits of such mechanisms, as well as the maximum efficiency that could be attained, have yet to be quantified in the context of a real life system. Having a solid idea about the trucking industry's maximum efficiency is
124 Logistics systems for sustainable cities important because it provides City Logistics advocates and planners with an indication of what could be achieved by efficiency enhancers, e.g. Internet based markets. The main objectives of this research are: (a) to develop methodologies that describe the relationship between dynamic supply and demand and the overall efficiency of the trucking industry; and (b) to provide estimates of the maximum efficiency attainable in the context of a case study. This is accomplished through the development of methodologies to estimate the dynamic vehicular supply as a function of time-varying commodity flows, vehicle utilisation patterns, fleet characteristics, and the corresponding level of efficiency. In doing so, this study contributes to the understanding of the dynamics of vehicular supply and its interaction with demand. The paper provides supporting information for policy makers and researchers about the upper bound on benefits that could be attained by market enhancing mechanisms. This paper summarises, expands, and reinterprets research originally conducted by Holguin-Veras (1984) that, for reasons that belong to another kind of publication, the author did not seek to publish. The paper is comprised of six sections. Dynamic vehicular supply model provides the conceptual and analytical framework used. Simulation results describes the test case and discusses the main results. The economic value of efficiency increases analyses the economic implications of changes in the percentages of empty trips. Implications for City Logistics provides a summary of the key lessons for City Logistics projects. Conclusions summarises the key findings.
DYNAMIC VEHICULAR SUPPLY MODEL The vehicular supply at a zone i is a function of two dynamic, time-varying, components: (a) the resident supply; and (b) the incremental supply (Holguin-Veras, 1984). Resident supply refers to the commercial vehicles that have their home bases in zone r, and are still awaiting shipments to transport. Incremental supply refers to the commercial vehicles that come to zone / to transport cargoes, stay in zone ;' for a while, participate in the local transportation market, and then either move back to their home bases or move on to another destination. The effective supply at zone ;' is the summation of the resident and the incremental supply for time period t. The unused vehicular supply at the end of the day is referred to as remaining supply. The effective supply has a significant impact upon the commercial vehicle choice process because it determines the time varying composition of the choice set and, ultimately, affects the market prices. The dynamic nature of the effective supply is depicted schematically in Figure 1 that shows the resident (unshaded rectangles), incremental (shaded rectangles) and the effective supply for a two zone system with a travel time between them equal to T. Each zone has a resident supply of five commercial vehicles at the beginning of the work day (t=0). At time t=0, it has been assumed that two truck loads are sent from zone 2 to zone 1. As a result, the resident supply is reduced from five to three vehicles at zone 2. At time t = T, the effective supply at zone 1 increases to seven vehicles (the five original residents plus the two vehicles that came from zone 2). At time t = T, one truck load departs from zone 1 to 2 reducing the resident supply at zone 1 from five to four. Two empty trips (return trips depicted as a dashed line) are also generated. At time t = 2T, the effective supply at zone 2 equals six. As implied by Figure 1, both components of the vehicular supply compete in a dynamic fashion for the transportation of commodities. For instance, at time t = T, the two units of incremental supply could have transported the goods from zone 1 to zone 2. Furthermore, had the incremental supply been able to transport the cargoes to zone 2, the net systemwide impact would have
On the estimation of the maximum efficiency of the trucking industry
125
been a reduction in the number of truck-trips equal to two trips (one loaded trip from 1 to 2, and the empty return).
Figure 1 Effective, resident and incremental vehicular supply Figure 1 highlights the fundamental premises of this research that: (a) the percentage of empty trips reflects the competition between the incremental and the resident supply; and (b) the percentage of empty trips—a measure of the overall efficiency of the trucking system—affects the total number of trips and consequently the vehicle-kilometres-travelled (VKT). As a result, policies that reduce empty trips have a significant impact upon traffic congestion and the externalities associated with freight activity. It follows that the outcome of this competition process determines the systems efficiency, which in this paper has a rather specific definition. If one considers that the output of the freight transportation process is the amount of trips with cargoes (loaded trips), it follows that a measure of efficiency is given by the ratio of loaded trips to the total number of trips made (loaded+empty), which is equal to the number one minus the percentage of empty trips. This simple metric is the one used in the paper. It should be noted that explicit modelling of the competition between incremental and resident supply, which could be done with dynamic game theory, is not the main focus of this investigation. The key focus is on the estimation of—for a given set of time varying commodity flow matrices and a percentage of empty trips—the segment of the truck supply that is the most likely to transport the cargoes. This assignment process is nothing more than an idealisation of the market competition process that takes place between the incremental and the resident supply. An important element to be taken into account is the trip purpose- This paper focuses on two key purposes: (a) delivery, D; and (b) pick-up, PJJ, that take place with probabilities P(D) and P(PU)- This assumption is supported by origin-destination surveys that indicate that these purposes account for 90-95% of the total trips (e.g., Holguin-Veras and Thorson, 2003). In the case of deliveries, it should be evident that if the incremental supply does not participate in the transportation of the goods, the percentage of empty trips, Pe, would be equal to 50%. At the other end of the spectrum, the maximum efficiency of the truck system is achieved when the
126 Logistics systems for sustainable cities use of the incremental supply at the destination of the primary trips is maximised (which implies minimising the usage of the resident supply and the minimisation of the percentage of empty trips). For modelling purposes, the assignment between resident and incremental supply is determined by the percentage of empty trips for the different types of vehicles. The probability of using a commercial vehicle from the resident supply is given by equation (1); while equation (2) represents the probability of using a unit from the incremental supply. (1) (2) Where Pe is the percentage of empty trips. Trips to pick-up cargoes entail commercial vehicles leaving their home bases empty to pick-up cargoes at the destination. The probability that a unit of the resident supply gets the shipment is given by equation (3); while equation (4) estimates the probability that the cargo is transported by the incremental supply. (3) (4)
Equations (1) to (4) can be combined to obtain the probability of using the resident supply: P(R) = P{D)P(RID) + P(PU)P(RIPU) (5) Substituting (1) and (3) for P(R/D) and P(R/PU) in equation (5), it follows that: (6) Since P(PU) = 1 - P(D), equation (6) can be rewritten as: p{R)J.S-PjPU)Pe
(7)
Equation (7) reduces to equation (1) if the probability of pick-ups is zero. If there are only pick-ups, P(PU) = 1, equation (7) becomes equation (3). Similarly, the probability of using the incremental supply is: P(I) = P(.D)PU/D) + P(PU)P(I/PU) (8) Substituting (2) and (4) for P(I/D) and P(I/PU) in equation (8) it follows that: (9) Since P(PU) = 1 - P(D), equation (9) can be rewritten: (10) If there are no deliveries, equation (10) becomes equation (4); if there are no pick-ups, equation (10) reduces to equation (2). It should be noted that it is quite possible to have, on a directional basis, percentages of empty trips larger than 50%. This usually occurs around major freight generators, e.g., ports, where most of the net freight traffic flows in one direction. This study simulates the vehicle flows in a way similar to that shown in Figure 1. The DYnamic VEhicular Supply Simulator (DYVESS) estimates the dynamic effective supply at the various transportation analysis zones by conducting a comprehensive simulation of truck trips on the basis of the time dependent origin-destination matrices by commodity type (using a
On the estimation of the maximum efficiency of the trucking industry
127
2 hour time period), the resident supply of commercial vehicles, a set of empirical vehicle choice probabilities, the observed percentages of empty trips, the travel time matrix, and the operational characteristics of the vehicles considered. DYVESS enables the user to specify the key input parameters, which provides it with considerable flexibility to adapt to multiple environments. Appendix A describes the simulation logic (after Holguin-Veras, 1984).
SIMULATION RESULTS This study simulated the truck freight transportation system in the Dominican Republic, which was selected as the case study because of the availability of two critical data sources: (a) a 1982 freight origin-destination survey with time-varying commodity flows; and (b) the Registry of Commercial Vehicles (RCV), an in-depth 1981 comprehensive registration of the vehicular supply in the country. Freight in the Dominican Republic is almost exclusively moved by trucks, with railroads, inland and coastal transportation moving a negligible amount. The freight origin-destination (OD) survey collected data about shipment sizes, origins and destinations (including time-varying OD data), and economic sectors involved in the transaction. The RCV, which required the mandatory registration of all commercial vehicles in the country, collected data about the physical and operational characteristics of commercial vehicles, as well as basic information about industry structure and managerial practices. Taken together, the RCV and the freight OD survey represent a unique data set because they provide, for the same time period, highly detailed information about both freight transportation supply and demand. The simulations were conducted using three different types of commercial vehicles (after Watanatada, 1987) that were further aggregations of the eleven original vehicle classes: Light vehicles (empty weight of less than 1,500 kg, or maximum payload capacity of less than 750 kg); Medium trucks (empty weight larger than 1,500 kg, or maximum payload larger than 750 kg, and gross weight limit less than 8.5 metric tons); and Heavy trucks (more than two axles and gross weight limits higher than 8.5 metric tons). The simulation system was successfully calibrated so that it replicated the total number of trips made by each of the vehicle types. Upon calibration, DYVESS was used to estimate the components of the dynamic vehicular supply (i.e., resident, incremental, effective and remaining) for the different time periods. The numbers of vehicles in the light truck class, i.e., pick up trucks, were estimated using the license plate registrations because the RCV did not require them to register. Four different scenarios of percentages of empty trips were considered. Scenario I used the percentages of empty trips found in the origin-destination sample. Scenarios II and HI represent two intermediate cases with percentages of empty trips reduced by 15% and 35%. Scenario IV (0% empty trips) was included to estimate the minimum value of the percentages of empty trips that could theoretically be achieved. This is possible because DYVESS attempts to produce an assignment of the supply consistent with the percentages of empty trips specified by the user. If this cannot be done, which happens when the incremental supply or the demand has been exhausted, DYVESS assigns the remaining trips to the resident supply, and re-computes the percentage of empty trips. This provides the lower bound on the percentage of empty trips. Table 1 shows the input and the attained percentages of empty trips, and the remaining supply for the different scenarios. "Test Pe" refers to the percentage of empty trips that was specified as an input. "Attained Pe" is the percentage of empty trips that can be actually achieved given the origin-destination patterns and the geographic distribution of the vehicular supply. As
128 Logistics systems for sustainable cities expected, the attained Pe for a given vehicle is larger or equal than the corresponding test Pe. The value of the attained Pe corresponding to a test Pe equal to zero represents the absolute minimum value of the percentage of empty trips that can be attained for that particular vehicle class.
Table 1 Remaining supply vs. percentage of empty trips
Scenario I (Base case) II III IV
„, „ TestPe 45% 30% 10% 0%
Light trucks Attained Remaining Pe supply 46% 8,137 32% 8,726 18% 9,090 12% 9,207
Medium trucks Attained Remaining Pe supply 45% 46% 3,877 30% 32% 4,630 10% 23% 4,905 0% 19% 4,988
„, _ TestPe
31.47 Unit change:
Unit change:
^ „ TestPe 45% 30% 10% 0%
Heavy trucks Attained Remaining Pe supply 45% 708 30% 849 16% 974 12% 988
41.15 Unit change:
8.48
The last line in Table 1 shows that a 1 % reduction in the percentage of empty trips translates into an increase in the remaining supply of 31.47 light trucks, 41.15 medium trucks, and 8.48 heavy trucks. The impacts of such an increase are summarised in Table 2. As shown, a 1% reduction in empty trips would produce a 2% reduction in the number of trips that would be able to absorb an increase in demand equal to 1.7% of the total demand in a typical day, measured by either tons or ton-km and, ultimately, it would reduce VKT by slightly more than
Table 2 Impacts of 1% reduction of empty trips Type of vehicle Light trucks Medium Heavy Total Country total % of total
remaining , supply 21 41 9
Trip length 6 , (km) 75 105 115
Average _ . , _ , , B, Trips saved Total tons vpayload 3 1.5 8.00 22.00
42 82 18 142 6,262 2.27%
Total ton- _ , , „ , _ . Total VKT kms
32 2,363 328 34,440 198 22,770 558 59,573 33,223 3,275,435 1.68% 1.82%
3,150 8,610 2,070 13,830 1,143,917 1.21%
Figure 2 shows in a graphical format, the relationship between the test Pe and the attained Pe for the three different vehicle classes. As shown, the numerical experiments determined that percentages of empty trips could be significantly reduced from the observed values (on average 45%) to the 12-19% range. These values would be achieved if and only if the level of market transparency increased by means of virtual markets, where shippers and truckers could "meet" on the Internet; or by the implementation of consolidation stations, where they physically meet. Figure 2 also indicates that the difference between the test Pe and the attained Pe increases when the test Pe gets smaller (which is particularly noticeable in the case of medium trucks). The author's conjecture is that this pattern is a reflection of increasing marginal costs of empty trip reductions. In other words, reducing the percentage of empty trips from 45% to 40% is easier than reducing it from 15% to 10%, which means that the marginal cost of reducing the empty trips is non-linear. This has important implications on the optimal value of empty trips. The pattern of Figure 2 hints at why percentages of empty trips lower than 30% are extremely rare. As shown, for percentage of empty trips larger than 30% there is almost no difference
On the estimation of the maximum efficiency of the trucking industry
129
between the Test Pe and the Attained Pe; while for the values less than 30% the difference between them grows larger. Assuming that the difference between these two values reflects the marginal costs of reducing empty trips, this would suggest that the trucking industry does not reach percentages of empty trips below 30% simply because it is too difficult to do so.
Test Pe
Figure 2 Attained vs. test percentages of empty trips In order to provide the reader with an idea about the key features of the dynamic supply, this section presents and analyses two different cases: (a) the base case, where empty trips occur with the percentages found in the sample; and (b) the most efficient system, i.e., the one in which the empty trips are minimised. DYVESS was used to construct estimates of the different components of the dynamic vehicular supply for the 2-hour time periods. Figure 3 shows the vehicular supply of medium size trucks for zone I (Santo Domingo) and the base case. As shown, the resident supply decreases with time of day as more trucks depart to their destinations. The incremental supply peaks around noon time and then decreases. More significantly, the incremental supply (448 units) represents a sizable portion (17%) of the effective supply (2,605 units) in zone I at noon time. This incremental supply may have a noticeable impact on the price structure because they would be able to bid at lower prices than the local truckers. Figure 4 shows the dynamic supply in zone I corresponding to the maximum efficiency. As shown in Figure 4, the downward slope of the effective supply—that is related to the utilisation of the trucks to transport the cargoes—is approximately half the value of the slope of Figure 3. In other words, the same amount of cargoes can be transported with one thousand fewer trucks, if the system is used at maximum efficiency. This shows that increasing the efficiency of the system has a significant impact on vehicle-kilometres-travelled and, ultimately, in air pollution emissions.
130 Logistics systems for sustainable cities
Figure 3 Dynamic supply of medium size trucks in Zone I (base case)
Figure 4 Dynamic supply of medium size trucks in Zone I (minimum % of empty trips)
THE ECONOMIC VALUE OF EFFICIENCY INCREASES This section provides estimates of the net economic impacts attributable to reductions in the empty trips, which are equated to increased systemwide efficiency. Although no estimates of the unit economic value of truck traffic's externalities exist for the Dominican Republic, an estimate could be obtained by using the unit costs that have been produced for American conditions. Among other things, this would help to get an idea about the relative order of magnitude of economic impacts. For the sake of convenience and because of its relative timeliness, the author decided to use the estimates put together by Forkenbrock (1999), that estimated the economic value of these externalities as 1.11 cent/ton-mile (0.76 cent/metric tonkm); and the corresponding private (operating) costs as 8.42 cent/ton-mile (5.79 cent/metric ton-km). On the basis of these assumptions, a 1% increase in the efficiency of the trucking system would produce a reduction in externalities with a economic value of approximately
On the estimation of the maximum efficiency of the trucking industry
131
$25,000/day ($8 million/year); while an increase in efficiency of 15% would translate into $375,000/day ($120 million/year) for the duration of the improvements. The savings to the operating costs are also significant. A 1% reduction of empty trips would represent approximately $190,000/day or $62 million/year; while a 15% reduction would translate into $2.84 million/day, or $924 million/year. It is also important to note that efficiency increases would invariably produce a reduction in shipping costs. This is because the marginal costs for members of the incremental supply (that already made a delivery and were planning to return empty), are much lower than the marginal costs of the members of the resident supply (because they have to make a loaded and an empty trip). This seems to indicate that shippers may be the big winners in this process, because they are the ones that may benefit from reduced fares. The results obtained in this paper also highlight the existence of an optimal value of the percentage of empty trips. The estimates of Table 2 indicate that the net economic benefits attributable to a reduction in empty trips may be assumed to be directly proportional to the reduction in Pe, which translates into a marginal benefit function constant and equal to mj. On the other hand, since the marginal costs associated with reducing empty trips are likely to increase, this would lead to an optimal value of the percentage of empty trips lying at the intersection of the marginal costs and marginal benefits functions. This indicates that the optimal value of the empty trip percentage, P* is higher than its minimum value, signalling that minimising empty trips is not a good idea because in doing so it would incur marginal costs so high that societal benefits would be overwhelmed. Instead, trying to optimise the amount of empty trips is a more rational decision. One aspect that may need further study is the impact of more efficient markets upon industry structure. An "optimised" system with the lowest possible value of the percentage of empty trips may lead to a trucking industry concentrated around the areas of the country that are major exporters of transportation services (because these companies could favourably compete with the local companies at the destination). Since increased efficiency leads to fewer trucks being used, there may be implications in terms of income distribution. The simulation results suggest that, as a consequence of the increased efficiency, the numbers of idle trucks at the end of the day (remaining supply) would experience sharper increases (in the range of 138% to 217%) in the poorest regions of the country. In contrast, idle trucks would increase 17% to 77% in the wealthiest regions. The extra pressure on local truckers may translate into further concentration of the trucking industry around the major economic poles. In spite of the concerns noted above, it is evident that a more transparent market would lead to a more efficient utilisation of the vehicular supply that, ultimately, may pave the way for significant reductions in the externalities produced by truck activity. The estimates produced here, in spite of its inherent limitations, clearly indicate the potential benefits attributable to market enhancement mechanisms (e.g., virtual markets, consolidation stations).
IMPLICATIONS FOR CITY LOGISTICS This research highlights a number of findings, relevant to the City Logistics movement, which provide guidance to public and private sectors on issues to be taken into account when putting together programs to reduce empty trips. Among them, it is important to highlight that:
132
Logistics systems for sustainable cities
1. The maximum efficiency of the trucking industry is constrained by the underlying commodity flows. Although significant reductions in empty trips are possible, their magnitudes are constrained by the availability of commodity in the backhaul direction. 2. Different truck types exhibit different upper bounds of efficiency. Since different truck types serve different demand segments, some types are more capable of reducing empty trips. This, together with the fact that the different truck types produce different amounts of externalities, suggests that City Logistics programs should target the truck types with the largest potential to reduce externalities. 3. It is not practically possible, nor economically sound, to eliminate empty trips. As shown by the simulations, the lowest percentages of empty trips are in the range of 1219%. Since the marginal costs associated with achieving such reductions are likely to increase, while the marginal benefits are constant, it follows that the economically optimal percentage of empty trips is much higher than the technically lowest values. 4. Increasing efficiencies may bring about distributional impacts. Since increased efficiency leads to fewer trucks being used, income distribution effects are a concern, hi the case study, the regions that experience the largest increase in non-utilised trucks are the poorest in the country. This raises the troubling prospect that increased efficiency may negatively impact those that are in most need of the income produced by trucks.
CONCLUSIONS The main objectives of this research were: (a) to analyse the relationship between dynamic supply and demand and the overall efficiency of the trucking industry; and (b) to provide estimates of the maximum efficiency attainable in the context of a case study. These objectives were accomplished by means of modelling the dynamics of vehicular supply and its interrelationship with the underlying commodity flows and empty trips. This research: (a) established that empty trips are the result of the competition process between the resident and the incremental supply; (b) related the probability of using either the resident or the incremental supply to a function of key variables; (c) defined and made operational dynamic relations between supply and demand in a simulation system; and (d) produced estimates of the economic impacts attributable to efficiency increases. In all, these contributions enhance the state of the art of freight transportation modelling by providing a more accurate depiction of the interplay between vehicular supply and demand. The simulation results indicated that a 1% reduction in the percentage of empty trips generates an increase in effective vehicular capacity equivalent to 31.47 light trucks, 41.15 medium trucks, and 8.48 heavy trucks. This translates into a 2% reduction in the number of trips, that would be able to absorb an increase in demand equal to 1.7% of the total freight demand in a typical day, measured by either tons or ton-km. hi terms of vehicle-km-travelled, it would reduce VKT by slightly more than 1%. The economic value of reductions in the amount of externalities is considerable. A 1% reduction in the percentage of empty trips would reduce the amount of externalities in an amount of approximately $8 million/year; while a reduction of 15% would translate into $120 million/year savings for the duration of the improvements. The savings in operating costs are also significant. A 1 % reduction of empty trips represents close to $62 million/year; while the 15% reduction discussed before would translate into $924 million/year savings.
On the estimation of the maximum efficiency of the trucking industry
133
The research conducted here outlines a number of implications for City Logistics programs aimed at reducing commercial vehicle empty trips. The first one is that the maximum efficiency of the trucking industry is constrained by the underlying commodity flows. The second one, which follows from the first, is that different types of vehicles have different lower bounds of efficiency. The third one, which is also a consequence of the first, is that reducing the percentage of empty trips is likely to face increasing marginal costs, hi other words, the lower the percentage of empty trips, the more expensive it becomes to reduce the empty trips even further (because of the increased difficulty of finding cargoes for the back-haul movement). The fourth one is a consequence of the third one: since the marginal benefit is expected to be constant and proportional to the reduction, it follows that there is an optimal value of empty trips which is higher than the absolute minimum. The fifth implication is related to the fact that efficiency increases may bring about distributional impacts that may negatively affect the smaller operators. This research has shed light on the economic and operational impacts of reducing the empty trips, concluding that reducing empty trips is both practically feasible and economically beneficial. Although the effectiveness of policies aimed at reducing empty trips was not analysed, it seems clear that policies and programs that increase the level of market transparency and the ability of shippers and truckers to "meet" have the most potential to achieve the kind of benefits estimated in this research. In spite of the contributions of this paper, the empirical and theoretical analyses of such policies and programs still remain a topic for future research.
ACKNOWLEDGEMENTS Support for the original research, upon which this paper is based, was provided by a fellowship from the Organisation of American States. The additional research was supported by grant CMS-0092938 from the National Science Foundation's CAREER Award program.
REFERENCES Holguin-Veras, J. (1984). Development of a Model for the Quantification of the Vehicular Supply in Freight Transportation, The Universidad Central de Venezuela, July 1984 (in Spanish). Holguin-Veras, J. (2000). A Framework for an Integrative Freight Market Simulation, Published by the IEEE 3rd Annual Intelligent Transportation Systems Conference ITSC-2000, Dearborn Michigan October 2000, pp. 476-481. Holguin-Veras, J. and E. Thorson (2000). An Investigation of the Relationships Between the Trip Length Distributions in Commodity-based and Trip-based Freight Demand Modelling, Transportation Research Record #\101 pp. 37-48, September 2000. Holguin-Veras, J. and E. Thorson. (2003). Modelling Commercial Vehicle Empty Trips with a First Order Trip Chain Model, Transportation Research Part B, Vol. 37 (2), pp. 129148. Holguin-Veras, J., (2002). Revealed Preference Analysis of the Commercial Vehicle Choice Process, Journal of Transportation Engineering, Vol. 128, No. 4, American Society of Civil Engineers, pp. 336-346 Forkenbrock, D. (1999). External Costs of Intercity Truck Freight Transportation, Transportation Research Part A, Vol. 33, pp. 505-526.
134
Logistics systems for sustainable cities
Watanatada, T. (1987). The Highway Design and Maintenance Standards Model, Baltimore; London: Published for the World Bank [by] the Johns Hopkins University Press.
APPENDIX A: DYVESS (DYNAMIC VEHICULAR SUPPLY SIMULATOR) 1. For a given time period t: 2. Randomly select a commodity type C: 3. Select a origin O: 4. Select a destination, D, randomly from the available set of destinations for commodity C (as a function of the amount of cargoes going to D): 5. Use the probabilities of choosing vehicle type V from origin O to destination D and commodity C, P = f(O, D, C), to estimate the amount of cargoes transported by vehicle type V (the probabilities are an input): 6. Estimate the number of vehicle-trips using average payloads: 7. Randomly assign the cargo to either the resident or the incremental supply: 8. Estimate the arrival times at destination D, T, for the different types of vehicles (the arrival times are estimated as a function of the travel distance, vehicle speed, loading/unloading and rest times): 9. Update the resident supply at origin O, time period t: 10. Update the incremental supply of vehicles V at destination D, time t + T: 11. If all cargoes departing from origin O have been simulated, proceed to analyse origin O+l, otherwise, go to 4: 12. If done with commodity C, randomly select commodity C+l; otherwise, go to 3: 13. If all cargoes originating at time period t have been accounted for, proceed to analyse t+1; otherwise, go to 2:
10
MODELLING EFFECTS OF E-COMMERCE ON URBAN FREIGHT TRANSPORT
Eiichi Taniguchi, Kyoto University, Kyoto, Japan Yasushi Kakimoto, Osaka City, Osaka, Japan
ABSTRACT This paper presents models of vehicle routing and scheduling with time windows and traffic simulation for evaluating the effects of e-commerce on urban freight transport and the environment. The models were applied to a test road network. Results indicate that introducing e-commerce (B2C) may lead to more traffic in urban areas and have negative impacts on the environment unless e-commerce is widely used by consumers to some extent. However, some measures including co-operative freight transport systems of home delivery companies, designating time windows by home delivery companies and pickup points are effective to reduce the total costs as well as total running times and NOx emissions.
INTRODUCTION Recently e-commerce has become popular with the rapid development and use of the internet for commerce. The amount of e-commerce in Japan is 1.5 trillion Japanese Yen in 2001 that increased 1.8 times of the previous year. It is estimated that e-commerce will reach 125 trillion Japanese Yen in 2006. There are various types of e-commerce including B2B (Business to Business), B2C (Business to Consumers) and C2C (Consumers to Consumers). For the traditional way of distribution, goods were transported from manufacturers to retailers via wholesalers, and consumers went to retail shops to purchase commodities. In contrast, using e-commerce for B2C, goods can be directly transported from manufacturers to consumers by delivery companies based on the order from consumers via the internet. Therefore, direct home delivery of goods by delivery companies will increase and person trips for shopping by private car will decrease with the deployment of e-commerce (B2C).
136 Logistics systems for sustainable cities Visser et al. (2001) discussed the effects of e-commerce on urban transport. They pointed out that e-commerce will stimulate home deliveries which will lead to less consolidated deliveries thus to more freight traffic. More freight traffic is not good from both environmental and commercial point of views. An important factor is that home delivery based on e-commerce often has time windows for delivery at each home. The designated time windows for home delivery will increase the urban freight transport for delivery trucks. However, less shopping traffic by car due to replacement of traditional shopping to e-shopping can reduce the increase of urban truck traffic to some extent. This paper endeavours to understand and evaluate the effects of e-commerce on urban transport from environmental and commercial point of views. We will develop mathematical models for simulating urban traffic on road networks and estimating the effects of ecommerce on urban transport. Thompson et al. (2001) presented a model for evaluating the effects of e-commerce on urban freight transport. They assumed constant travel time for a link. However, we explicitly incorporated the change of travel times of links by dynamic traffic simulation. Taniguchi et al. (2001) proposed several city logistics measures for establishing efficient and environmentally friendly urban distribution systems. We will examine the effects of a number of city logistics measures including co-operative freight transport systems and relaxing the strict time windows by delivery companies, for decreasing the negative impacts of e-commerce on the environment.
PROBABILISTIC VEHICLE ROUTING AND SCHEDULING MODEL Outline of model This study adopts a probabilistic model for Vehicle Routing and scheduling Problems with Time Windows - Probabilistic (VRP-TW-P) (Taniguchi and Thompson, 2002) for evaluating the effects of e-commerce on urban transport. This model represents the behaviour of freight carriers who have their own depot and multiple pickup/delivery trucks that depart from the depot and visit several customers for collecting or delivering goods to return to the same depot. For developing the models we assume: a) The size, number and capacity of pickup/delivery trucks at the depot is known b) The location of customers is known and the demands of customers are given c) Only collection or delivery of goods will be carried out in an operation of a truck for visiting customers d) Each customer specifies a designated time window to be visited by a pickup/delivery truck Figure 1 shows the modelling framework. The VRP-TW-P model takes into account the uncertainty of travel times on the road network. This model intends to identify the optimal routing and scheduling for each delivery company. The distribution of travel times on each link of the network is determined by traffic simulation model (BOX model) which will be described later. As Figure 1 indicates, updated travel times on each link will be used in the VRP-TW-P model again. Thus the iterative procedure will give a certain solution. As this procedure cannot guarantee the conversion, we need to determine a stopping criterion.
Modelling effects of E-commerce on urban freight transport
137
Figure 1 Model framework
Formulation of model The VRP-TW-P model determines the optimal solution by minimising total costs. The total costs are composed of three components; (a) the fixed cost of vehicles, (b) vehicle operating costs that are proportional to the time travelled, and (c) early arrival and delay penalties for designated pickup/delivery time at customers. The model can be formulated as follows: Minimise (1) where, (2) (3) Subject to (4)
138
Logistics systems for sustainable cities
(5) (6) (7) (8) (9) (10) where t',,0 =t,,0 + 'Z{T(IIMi)Mi),n(i
+ l)) + tcMM)}
(11)
j=0
C(to,X): total cost (yen) t0: departure time vector for all vehicles at the depot
'o = ko|' = 1 . m] X: assignment and order of visiting customers for all vehicles X={x,|/=l,m} x,: assignment and order of visiting customers for vehicle /
x,={n(01/ = !,#,} n(i) : node number of i th customer visited by a vehicle d(j): number of depot (= 0) N,'. total number of customers visited by vehicle / n0: total number of d(j) in x; m : maximum number of vehicles available cf,: fixed cost for vehicle / (yen /vehicle) S, (x ; ): = 1; if vehicle I is used, = 0; otherwise C,, (t,0 , x,): operating cost for vehicle / (yen) CpJ (tl0, x ; ): penalty cost for vehicle / (yen) ct /: operating cost per minute for vehicle / (yen /min) tt n(0 : departure time of vehicle / at customer n(i) T(tlnU>,n(i),n(i + Yf) • average travel time of vehicle / between customer n{i) and nii + l) at time
',.„(,) tc n ( 0 : loading/unloading time at customer nil) />(.»(» ('/,o>r'xi): probability in which a vehicle that departs the depots at time t,0arrives at customer n(i) at time t cd n(j) {t): delay penalty cost per minute at customer n{i) (yen/min) ce n(j) (i) '• early arrival penalty cost per minute at customer n(i) (yen/min) TV : total number of customers
Modelling effects of E-commerce on urban freight transport
139
D(n(O): demand of customer n{i) (kg) t\ „ : last arrival time of vehicle / at the depot ts : starting of possible operation time of trucks te: end of possible operation time of trucks W, (x ; ): load of vehicle / (kg) Wcl: capacity of vehicle / (kg). The problem specified by equations (1) - (11) involves determining the variable X, that is, the assignment of vehicles and the visiting order of customers and the variable t0, the departure time of vehicles from the depot. Note, that n (0) and n (N, +1) represent the depot in equations (2) and (3). Figure 2 shows the penalty for vehicle delay and early arrivals at customers. The time period (tenM-t*u)) of the penalty function defines the width of the soft time window in which vehicles are requested to arrive at customers. If a vehicle arrives at a customer earlier than tsaU), it must wait until the start of the designated time window and a cost is incurred during waiting. If a vehicle is delayed, it must pay a penalty proportional to the amount of time it was delayed. This type of penalty is typically observed in goods distribution to shops and supermarkets in urban areas. Multiplying the penalty function and the probability of arrival time, the penalty for early arrivals and delays at customers for the probabilistic model can be estimated (Figure 2). The problem described here is a NP-hard (Non-deterministic Polynomial-hard) combinatorial optimisation problem. It requires heuristic methods to efficiently obtain a good solution. Recently several researchers have applied heuristic algorithms such as Genetic Algorithms (GA), Simulated Annealing (SA) and Tabu Search (TS) to obtain approximate solutions for NP-hard combinatorial optimisation problem. The model described in this paper uses a GA to solve the VRP-TW-P. GA was selected because it is a heuristic procedure that can simultaneously determine the departure time and the assignment of vehicles as well as the visiting order of customers.
140 Logistics systems for sustainable cities
Figure 2 Penalty of early arrival and delay at customers for the probabilistic model
DYNAMIC TRAFFIC SIMULATION MODEL The dynamic traffic simulation model is based on the BOX model that was originally developed by Fujii et ah (1994). The modified BOX model explicitly describes the flow of pickup/delivery trucks that depart from a depot and return to the same depot. Pickup/delivery trucks are converted to passenger car units and the first-in-first-out rule is assumed on all links. The model was further modified to identify the arrival of specific vehicles at assigned nodes (customers). The travel times on each link vary within the day. The output of the BOX model is the updated distribution of travel times on each link. The BOX model calculates the distribution of travel times based on link travel times. The distribution of travel times representing the interval of one hour was formulated using data with uniform travel times over four-hour periods. For example, the distribution of travel times representing the time interval, 8:00-9:00 a.m. was determined using data between 6:30 - 10:30 a.m. Then this empirical distribution was used in the VRP-TW-P model.
Modelling effects of E-commerce on urban freight transport
141
CASE STUDIES Test conditions Figure 3 shows test road network that was used for the case studies. The main network has 25 nodes and 80 links. Each node of the main network has a sub network consisting of 9 nodes and 24 links. All nodes in the main network represent centroids that generate and attract passenger cars. It is assumed that retail shops are located at every node in the main network and that residents of sub network go shopping to that retail shop. There are two types of areas on main network, high population density areas (nodes 7-19 in Figure 3) and low population density areas (other nodes in Figure 3). High population density areas contain 130-170 households per node and low population density areas contain 80-120 households per node. The main network has 26,550 households in total. Consumers also live at nodes where depots are located. In the case studies, it was assumed that two types of commerce exist, traditional shopping at retail shops and e-commerce (B2C). The penetration rate for e-commerce was set at various levels, 0, 3, 5, 10, 20 and 50%. Traditional retailing requires that freight carriers transport goods from their depot to retail shops and people go shopping to the retail shop by either car or bicycle or on foot. In contrast e-commerce (B2C) requires direct home delivery from the depot to consumers homes by the home delivery company.
142 Logistics systems for sustainable cities
Figure 3 Test road network The following conditions were assumed for the case studies. (a) Retail shop A retail shop is located in the central node of sub-network and consumers within the subnetwork only shop at the retail shop in the same sub-network. (b) Freight carriers who deliver goods from the depot to retail shops There are three freight carriers who deliver goods from depot to retail shops. Their depots are shown in Figure 3. Each freight carrier can use 12 trucks at maximum with the capacity of 10 ton. Each freight carrier needs to keep same time windows of retail shops. (c) Home delivery companies who deliver goods from the depot to consumers Each home delivery company can use 50 trucks at most with the capacity of 2 tons and allocate to 2 trucks to each sub-network. The time windows of customers are randomly determined for 1 hour between 8 a.m. and 10 p.m. The effects of e-commerce were estimated for the following 9 cases: Case 1: Base case; consumer's needs = 5 kg; Consumers' time windows are randomly set between 8 a.m. and 10 p.m. for 1 hour; E-commerce users do not go shopping to the
Modelling effects of E-commerce on urban freight transport
143
retail shop; A single home delivery company; Home delivery trucks only visit consumers within one sub-network Case 2: Consumer's needs are increased to 8 kg Case 3: 50% of e-commerce users also go shopping to the retail shop Case 4: 3 home delivery companies can operate within network Case 5: Consumer' time windows are designated by the home delivery companies Case 6: Depot of home delivery company is located in the centre of main network Case 7: Home delivery trucks can visit consumers in two sub-networks Case 8: Pickup points where consumers come to pick up their goods is facilitated in the centre of sub-networks Case 9: 3 home delivery companies accept co-operative delivery systems (For Cases 2-9 other conditions are the same as Case 1) Case 1 provides the base case and Cases 2-4 represent changes in the life style of consumers and new home delivery companies come into the market. Cases 5- 9 are cases in which some measures are taken by home delivery companies to cope with their problems of increasing total costs.
Test results Table 1 shows the total costs of 9 cases with the change of penetration rate of e-commerce of 0, 3, 5, 10, 20 and 50%. Total costs are the sum of costs for freight carriers for transporting goods to retail shops and costs for home delivery companies for delivering goods to consumers. Note that costs include fixed cost, operation cost and early arrival and delay penalties. The total costs sharply increase when the penetration rate of e-commerce increases from 0% to 3%. This is attributed to the increase of costs for home delivery. The increase of total costs in Case 4 was caused by the increase of number of trucks used by the 3 home delivery companies separately. For Cases 4-8 and Case 9, the total costs were reduced by implementing measures compared with Case 1 and Case 4, respectively. This indicates that these measures are effective for reducing total costs in cases where e-commerce is used, hi particular, the co-operative delivery system was effective by significantly decreasing total costs.
144 Logistics systems for sustainable cities Table 1 Change of total costs Case 1 2 3 4
5 6 7 8 9
Penetration rate of e-commerce (%) 0
3
5
10
20
50
313.60 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
873.40 1.77 -2.97 122.35 -28.55 -1.29 -3.59 -28.43 -56.16
854.00 2.94 -0.74 127.52 -24.92 1.27 -6.81 -23.38 -57.82
826.50 0.95 2.22 136.56 -22.78 -1.82 -3.27 -21.90 -55.74
821.90 0.03 -2.12 128.36 -24.36 -1.56 -12.62 -23.84 -57.83
824.70 -3.67 0.72 126.58 -4.85 -6.53 -5.37 -10.92 -56.05
Case 1: total costs (thousand yen / day) Case 2-8: percentage to Case 1 Case 9: percentage to Case 4 Table 2 Change of total running times (excluding waiting time)
Case 1: total running times (hours/ day) Case 2-8: percentage change from Case 1 Case 9: percentage change from Case 4 Table 2 presents a comparison of total running times (excluding waiting times). The total running times include running times of passenger cars and trucks of freight carriers for transporting goods to retail shops and home delivery companies. For Case 1, the total running times increased slightly with the increase of penetration rate of e-commerce of up to 5% and then decreased with the increase of penetration rate over 10%. The reason for this may be that when the penetration rate increased over 10%, the reduction of traffic for shopping by passenger cars exceeds the increase from home delivery truck traffic. For Cases 4-8 and Case 9 the total running times decreased due to the decrease in running times of home delivery trucks. This indicates that some measures taken here are effective in reducing negative impacts on the traffic conditions and the environment. The pickup points as well as the cooperation of home delivery companies can substantially reduce total running times.
Modelling effects of E-commerce on urban freight transport
145
Table 3 Change of NOx emissions
Case 1: NOx emissions (g/ day) Case 2-8: percentage to Case 1 Case 9: percentage to Case 4 Table 3 allows NOx emissions to be compared between cases. NOx emissions represent the sum of NOx emissions of passenger cars either for shopping or other purposes and trucks for freight carriers for transporting goods to retail shops and home delivery companies. In Case 1, NOx emissions increased when the penetration rate of e-commerce increased to 3%, 5% and 10% compared with 0% and went down when the penetration rate of e-commerce became 20% and 50%. This result is similar to the result of total running times. It is recognised that there is possibility of generating negative impacts on the environment in terms of NOx emissions unless e-commerce becomes popular with over 20% of consumers. In Cases 3 and 4, NOx emissions increased compared with Case 1. The reason is that NOx emissions by passenger cars for shopping were added to Case 3 compared with Case 1 and that the number of trucks of home delivery companies increased with the increase of the number of home delivery companies. In Cases 5-8, the reduction of NOx emissions by home delivery trucks contributed to reduce the total NOx emissions. Therefore, measures taken for Case 5-8 are effective for improving the environment, and are similar to total costs and running times.
CONCLUSIONS This paper evaluated the effects of e-commerce on urban freight transport using modelling technique based on the vehicle routing and scheduling problem with time windows. Applications of models to a test road network indicated that introducing e-commerce (B2C) may lead to more traffic in urban areas and have negative impacts on the environment unless e-commerce is widely used by consumers to some extent. However, some measures including co-operative freight transport systems of home delivery companies and designating time windows by home delivery companies and pickup points are effective in reducing total costs as well as total running times and NOx emissions. In further studies, these models should be applied to more realistic situations and larger road networks. For example, there is a need to consider some measures taken by freight carriers for transporting goods to retail shops. It will be necessary to take into account consumers'
146 Logistics systems for sustainable cities preference in buying certain commodities using e-commerce.
REFERENCES Fujii, S., Y. Iida and T. Uchida (1994). Dynamic simulation to evaluate vehicle navigation. Vehicle Navigation & Information Systems Conference Proceedings, 239-244. Taniguchi, E., R.G. Thompson, T. Yamada and R. van Duin (2001). City Logistics - Network modelling and Intelligent Transport Systems. Pergamon, Oxford. Taniguchi, E. and R.G. Thompson (2002). Modelling city logistics, Transportation Research Record, Journal of Transportation Research Board, No.1790, 45-51. Thompson, R.G., C. Chiang and M. Jeebaptsa (2001). Modelling the effects of e-commerce, In E. Taniguchi and R.G. Thompson (eds.) City Logistics II, Institute of Systems Science Research, Kyoto. Visser, J., T. Nemoto and J. Boerkamp (2001). E-commerce and city logistics. In E. Taniguchi and R.G. Thompson (eds.) City logistics II, Institute of Systems Science Research, Kyoto, 35-66.
11
LAST-MILE, A PROCEDURE TO SET-UP AN OPTIMIZED DELIVERY SCHEME
Gaetano Fusco, Dip. Idraulica, Trasporti e Strode, Universita di Roma "La Sapienza ", Italy Luigi Tatarelli, Dip. Scienze dell Ingegneria Civile, Universita di "Roma Tre ", Italy Maria Pia Valentini, ENEA - Ene/Tec, Centro Ricerche Casaccia - Roma, Italy
ABSTRACT This paper illustrates a comprehensive procedure for the design of a delivery scheme in an urban area, where a set of possible locations for logistic transit-points exists. In such a scheme it is assumed that deliveries for Business-to-Consumer (BtoC) e-commerce of goods are performed at specific drop-points, suitably selected to match the pick-up of the parcels to usual activities of the customers, like the breakfast at the bar or the purchase of the newspaper. The solution procedure, which integrates transportation system theory and operational research techniques, applies a disaggregate Nested Logit Model (NLM) for demand estimation, an Analytic Hierarchy Process (AHP) to compare and select possible drop-points, and a double string Genetic Algorithm (GA) to solve jointly both the problems of transit-point location and sizing and of drop-point clustering for deliveries tours. A GA performance function is computed by solving a standard Travel Salesman Problem (TSP) on the road graph, whose travel times have been estimated by assigning the O/D matrix of car trips. The first application of the procedure to the town of Terni, in Italy, has provided very encouraging results.
INTRODUCTION In the Internet age, it is expected that the growth of the Business-to-Consumer (BtoC) ecommerce will result in more frequent shipments of small loads of goods to satisfy increasing consumer needs, with a negative impact on the urban mobility (increase of congestion) and the environment (noise and air pollution). Recent studies reveal that 10% of USA households bought on-line in 2000 and it is expected this will rise to approximately 50% in 2004 (Golob
148 Logistics systems for sustainable cities and Regan, 2001). It is also forecasted that revenue from e-commerce will be 6.3% of the whole commercial business in Europe. On the other hand, several studies have pointed out that urban good delivery processes are now characterised by very low load factors, a large number of direct trace runs, a wide use of illegal parking and old vehicle fleets. As a consequence of that, even though freight traffic is in general less than 15% of the total urban vehicular flow (STA, 2000; Ambrosini and Routier, 2000), about 30% of the road occupancy (in vehiclexkm) is related to goods movement (PatierMarque, 2000). To face these problems many projects have been undertaken all over the world in recent years. They include reinforcing traffic regulations for truck traffic or improving the efficiency of the distribution system by developing new facilities and applying new technologies (FHWA, 2000; Dablanc, 1998; Stough, 2001 and Visser et al., 1999). It is expected that the rapid growth of e-commerce, even if less impressive than many predictions, will increase these problems. In detail, home deliveries will become more prevalent; more frequent deliveries and even smaller loads are also expected. On the other hand, e-commerce is an important opportunity that can be exploited to optimise the pick up/delivery process and so reduce the total number of trips (persons and freight). This may lead to envision new more efficient delivery schemes, considering about 10% of the average cost of each e-commerce purchase is due to the final delivery cost (Picard, 2000). The study described in this paper arises from an agreement between the Italian Ministry of Environment, ENEA and the University of Roma "La Sapienza". The aim of the research was to develop a general procedure for the design and environmental assessment of logistic systems involved in the "last mile" management and apply it in pilot projects. Among these projects an innovative scheme for e-commerce deliveries in the town of Terni, in Italy, has been included. According to Ministry of Environment objectives, the new scheme should consider both the distribution costs, viewed from the perspective of private enterprises as well as the social and environmental costs, viewed from the perspective of the community. Reducing both costs can be achieved by making the whole system more efficient, for example by increasing load factors and delivery productivity as well as by improving routing and scheduling strategies. The latest urban transport master plan of Terni (Municipality of Terni, 2001) provides a new logistic transit-point along the main transport facilities crossing the town. Starting from this, ENEA and the University "La Sapienza" envisioned an innovative organisation for home deliveries based on the logistic transit-point and a set of drop-points spread across the urban area. The procedure proposed in the paper is essentially composed of two parts, one is finalised to estimate the demand levels and identify a good set of drop-points for delivering and picking, the other aims at determining the best localisation, size and operations of logistic transit-points among a given set of possible locations. The first problem is approached by applying a Nested Logit Model and the Analytical Hierarchical Process. The second problem is based on a Genetic Algorithm. The application of Genetic Algorithms to transit-point location is not new. Taniguchi et al. (1999), proposed a model for transit-point location starting from an O/D matrix of merchandise that was assigned to the road network complying the user equilibrium condition with variable demand. This paper is focused on e-commerce deliveries where the routing problem is included into the optimisation procedure, although without considering the network equilibrium (due to the low incidence of the delivering process on the total traffic
Last-mile, a procedure to set-up an optimized delivery scheme
149
volumes). Thus, we introduce a double string Genetic Algorithm that solves jointly the problems of transit-point location/sizing and drop-point clustering for delivery tours. Within each GA run, optimal tours are then determined by means of a standard TSP algorithm. A complete overview of the state of the art of the combined location-routing problem until 1998 can be found in Min et al. (1998). More recent work on location-routing problems are presented in Wu et al. (2002) and Lin et al. (2002). Readers interested in the problem of facility location can refer, for example to Owen and Daskin (1998).
PROBLEM DEFINITION AND ASSUMPTIONS The logistic scheme proposed here is based on the presence of two fundamental elements: a) a set of city transit-points, terminals where long-haul shipments are terminated and the urban distribution process is enhanced by means of load consolidation, implementation of advanced information systems and co-operative operations of freight deliveries; b) a set of drop-points, where the goods are delivered and consumers pickup their parcels. In order to avoid new car trips and parking needs as well as to minimise the customers discomfort when picking-up parcels, the guide idea is to join the pickup and a regular or enjoyable activity of the consumer (like having breakfast at the bar). Thus, commercial and public sites such as bars, postal offices, newspaper shops and bookstores have been selected as possible drop-points of the delivery scheme. Home deliveries are reserved for disabled people and for heavy goods. Nevertheless, customers can choose to receive home delivery and pay the relative charges. The design of the above logistic scheme consists of individuating the optimal combination of drop-points, transit-points and vehicular typology accordingly to given criteria. The main inputs of the problem are socio-economical variables of demand zones, the Origin/Destination matrix of car trips, the road network graph, the delivery fleet typology, the set of candidate areas for locating the transit-points and the set of possible drop-points. Thus, it can be formulated as a combined location-routing problem, which also requires a procedure for estimating the demand for BtoC good purchases. The optimal solution strongly depends on the objectives appointed for the design of the logistic scheme. In general, the delivery scheme involves many actors, which can be grouped into the following categories: • e-commerce customers; • transit-points operators; • urban delivery operators; • long-haul operators (both shippers and carriers); • urban community. Each of these groups have different goals from the other groups. Customers aim at minimising discomfort and costs to reach the drop-points and to bring parcels home. Transit-point operators seek to lower construction, maintenance and operational costs and higher revenues. Delivery and long-haul operators, for a given demand, want to minimise variable and fixed transportation costs. Last but not least, the urban community wishes to reduce the environmental and social costs of road traffic and the space required for deliveries.
150 Logistics systems for sustainable cities In conclusion, this is a complex multicriteria problem, in which a number of different and sometime conflicting features play a role. Nevertheless, it can be greatly simplified by introducing some plausible assumptions, summarised below: • each long-haul operator delivers the goods only to one transit-point (of his own choice); if necessary, the goods are then distributed among all the transit-points at night-time; • the deliveries are operated by one firm, which also operates all the transit-points; • only one vehicular typology is adopted for the transportation of goods from the transitpoints to the drop-point; if particular restrictions or conditions occur (such as pedestrian zone where only zero-emissions vehicles are admitted), so that different vehicular typologies have to be considered, then the problem is subdivided (and a sub-optimum can be reached); • the delivery process does not involve co-ordination among transit-points; in other words, trips start and end at the same transit-point without intermediate stops at other transit-points; • small (at least one) difference exists in the grounds' values for transit-point locations.
SOLUTION PROCEDURE General framework The main steps of the proposed procedure are summarised in the flow-chart depicted in Figure 1 and described in the following paragraphs. As results, the solution procedure determines: a) drop-points selection; b) size and location of city transit-points; c) size and composition of the vehicle fleet; d) optimal routes from each transit-point to each drop-point.
Figure 1 Design procedure framework
Last-mile, a procedure to set-up an optimized delivery scheme
151
Input data The procedure needs the following data relating to the study area: a) territorial partition into small zones for which demand data (like number of inhabitants by age, number of families, etc.) can be obtained; b) topology and functional characteristics of the road network for both driving and walking; c) traffic conditions throughout a standard working day; d) location and capacity of possible logistic terminals (transit-points); e) location and typology of possible drop-points for e-commerce deliveries; f) driver shift duration. In order to estimate the values of the objective function, economic and environmental parameters are also needed. These include: a) investment unit costs (per storehouse volume unit) to build the bases (these unit costs could be different by base location and/or typology); b) vehicle capacities as well as operation and buy-out costs; c) vehicle energy consumption and pollutant emissions as well as related unit external costs. In order to ensure possible solutions, it is necessary to check first if the average daily demand of any drop-point exceeds the capacity of vehicles and also if the total logistic transit-point capacity is sufficient for the average total demand.
Network description The road network is represented by a directed graph having L links and N nodes. Each link is characterised by its length, free-flow travel time and a speed-flow relationship. Link costs for freight vehicles are computed by performing a user equilibrium assignment of car traffic on the network. Demand estimation The study area is divided into traffic zones, which originate and attract good deliveries. Demand estimation is performed by applying a two step-procedure, which integrates aggregate macro estimations, which determine the total amount of demand attracted within the study area, and dis-aggregate models that calculate its distribution among traffic zones. Aggregate estimations were obtained from the technical literature. However, survey data was required to calibrate the dis-aggregate model in the form of nested logit model (Ben-Akiva and Lerman, 1985).
Drop-point typologies comparison Many public places spread throughout an urban area are visited daily by much of the active population. These places could be suitable to deliver small parcels without requiring additional distance to be travelled by customers. These include, for instance, cafes, newsagents, tobacconists, post offices, and so on. Many of these public places are often located very close to each other, thus a conflict could exist in choosing the best one for delivering. In principle, a detailed comparison could be made by evaluating the performance of each place according to pre-selected goals such as:
152 Logistics systems for sustainable cities • minimise the energy consumed for delivering and picking-up parcels; • minimise customer discomfort. The performance of drop-points depends on characteristics such as opening hours, attending frequency, storehouse capability, personnel activities and tasks, etc. In most cases evaluators do not know characteristics of each place but, as most of them are common to the commercial category, a comparison can be done by typology. The Analytic Hierarchy Process (AHP) is very suitable for such a task, since it does not require quantitative data. Results of AHP can be used in all cases when a conflict arises in drop-point selection in the study area. An example of AHP among drop-point typologies is shown in Figure 2. Such a scheme has also been used in the case study of Temi. Results of the analysis are be described below.
Drop-points selection This heuristic procedure for drop points selection is based on GIS functions and transport networks analysis. GIS contains information about: a) location and storehouse size of any possible drop-point (of pre-selected typologies); b) demographic (demand) data on the whole study area for each small zone (census zones).
Figure 2 Application of the Analytic Hierarchical Process to the drop-point analysis and comparison For each potential drop-point the influence demand area (i.e. the territorial area that could be served by that drop-point) is determined by progressive steps in the following way. 1. Verify, a) if the daily potential demand of the zone which the potential drop-point belongs to is compatible with the storehouse capacity, and, b) if the distance (along the shortest path) between the DP and the zone centroid is less than a pre-determined quantity (for instance 500 m) compatible with a pedestrian trip. 2. Enlarge the area including all neighbouring zones and repeat checks on capacity and distances; repeat until limits are satisfied.
Last-mile, a procedure to set-up an optimized delivery scheme
153
3. When a zone does not fit within the maximal distance, that zone is excluded from the influence area and the enlargement starting from it stops. 4. If distance limits are met, but the capacity is exceeded, then: - if the overflow is less then a given percentage „ of the drop-points capacity (e.g. 15%), then the influence area corresponds to the one found at the latest step, - else, it corresponds to that found at the (n-l)th step. 5. When an influence area overlaps the neighbouring ones, in general for more than a given percentage „ (e.g. 80%), the correspondent drop point is eliminated and the overlapping area is reallocated to the neighbour ones, each for its own competence. Similarly, when an influence area overlaps another one for more than a given percentage „ (e.g. 60%) of its area, then a process of pair elimination between the drop-points corresponding to these two areas has to be carried out: the drop-point obtaining the lowest value of the product (100-Oa)*P (where Oa is the overlapped area percentage and P is the score obtained in the previous AHP) is eliminated and its overlapping area is reallocated to the neighbouring areas, each for its own competence. Drop-point pair eliminations have to be made one by one, starting from the largest overlap. 6. Once distance limits are met and all major overlaps have been excluded, then a check on territorial cover has to be done If some area is not covered by drop-point influences, then some of the excluded drop-points have to be reconsidered. If some area is still uncovered, then it should be excluded from the delivery service or new drop-point typologies have to be taken into consideration. It should be noted that overlaps among influence zones are admitted (if not exceeding a certain amount). Thus, after the selection procedure is performed, a re-allocation of overlapped zones is requested. Vehicle choice Within the transit-point selection and size determination algorithm, the typology of the vehicle used for the deliveries is assumed as an input. Thus, a preliminary choice has to be made considering the following contextual elements: • geometrical road network characteristics, such as width and slope; • law and local regulations (on vehicle size, pollutant emissions, etc.) • delivery weight and size; • demand distribution all over the study area; • loading and unloading needs; • operational and buy-out costs; • driver comfort. Like for drop-points, a multi-criteria analysis could help in the selection of the best vehicular type among those on the market. An example of such a procedure is shown in the paragraph of the results for the case study. Transit-point selection and sizing Once drop-point locations, demand and vehicle types have been defined, the optimal set of logistic transit-points has to be carried out. The solution procedure searches simultaneously for both optimal facilities location and routing. In other words, we look for the solution that minimizes the amount of investment and operating costs, of both transit-points and vehicles, and external costs due to pollution from vehicles. Not only delivery transport but also feeding to transit-points from external zones should be considered. A general solution is represented in
154 Logistics systems for sustainable cities Figure 3. Here, each drop-point is associated with one transit-point and one delivery tour (obviously one transit-point can be associated with more than one tour). Such a solution is feasible if following constraints are kept: • tour demand amounts do not exceed vehicle capacity; • tour time does not exceed driver shift duration; • total daily demand of the drop-point set associated to one transit-point does not exceed the transit-point capacity. We assume that capacity of a transit-point is not a constraint, but must accommodate at least the largest drop-point goods. Thus, the distribution system cost is affected by: • size and location of selected transit-points; • cluster of drop-points; • path on the cluster.
Figure 3 Example of drop-point assignment to selected transit-point and construction of delivery tours After the drop-point/transit-point/tour associations have been defined and checked, the optimal course between one transit-point and each drop-point belonging to its set can be determined by solving the Travel Salesman Problem (TSP), considering real traffic conditions on the road network. Now all the elements (transit-point sub-set and related demand amount, distance covered, speed and travel time) exist to compute the value of the disutility function and compare the solutions against each other. The core of the problem, of course, is to find out the set of solutions. As the problem is NP-hard, exact algorithms are feasible only for small/medium sized problems; as a consequence, heuristic procedures have to be used. To solve this non-linear problem we propose here a double-string genetic algorithm that considers possible tours between drop-points and then searches for sub-optimal facility location and route design simultaneously. If M is the number of drop-points and the set of drop-points is sorted, the generic individual of the current population is then represented by a 2 x M matrix, where: — the i-th column is related to the i-th drop point; - the first row contains the allocation of drop-point to potential transit-points (the cell value is the transit-point code);
Last-mile, a procedure to set-up an optimized delivery scheme
155
the second row contains the allocation of drop-point to tour (the cell value is tour code). With reference to the network of Figure 3, the double string chromosome definition (Table 1). Table 1 Double String Chromosome Drop-point # Transit-point #
Tourtt
1
2
3
4
5
6
7
8
9
11 2
14 1
11 3
14 1
14 1
11 3
14 1
14 1
11 2
It is worth noting that the header row is unnecessary and has only been included here to aid understanding. The results from the algorithm are transit-point selection and sizing, drop-point clustering, tours and number of necessary delivery vehicles, as well as the amount of all costs considered within the performance function. A concise flow chart of the genetic algorithm is shown in Figure 4.
Figure 4 Flow chart of the genetic algorithm applied for transit-points and drop points matching
156 Logistics systems for sustainable cities Costs and impacts estimation The present procedure, for logistic systems design, requires the computation of investment, operational and external costs in order to allow a comparison among different solutions and find-out a good one. In order to estimate the investment costs for constructing the logistic bases, unit costs (per storehouse unit area/volume) have to be fixed, according to the market values of grounds in the study area. Similarly, fleet buy-out costs and operative vehicle costs can be computed once the investment per vehicle and an average operative cost per kilometre have been fixed. Personnel costs can be estimated using a given value of the hourly wage. Environmental costs were derived from pollutant emissions estimation by means, for instance, of COPERT functions, starting from vehicle typology, distance covered and average speed. Instead of costs per unit emission, cost per unit mileage can be used to determine the external costs of air pollution, so that emissions estimation can be avoided. Other costs to consider are related to drop-points. Actually, drop-points do not require any construction investment as, according to the scheme proposed here, they exploit existing facilities, which currently perform functions other than parcel picking. In spite of that, a cost for the new service supplied has to be provided, in terms of whether fare per delivered parcel or monthly wage. Such a cost, as well as the other cost of the delivery service, should be put down to the logistic operators in return of reductions in operation costs thanks to last-mile cost savings.
Check of consistency and feedbacks The solution procedure computes the location and size of transit-points, drop-point clusters and optimal delivery routes simultaneously, for a given demand pattern and an assumed vehicle fleet. Thus, once a solution has been obtained, a feedback should be carried out to verify if different hypotheses on fleet size and composition could improve the solution found. As a matter of fact, the number of vehicles and then the number of drivers greatly affects the total cost of the system.
APPLICATION TO THE TOWN OF TERM Project main features The idea to carry out an analysis on e-commerce delivery in the area of Terni came out from the Provincial Chamber of Commerce in 1999, which also adopted a new portal, for virtual trade of Umbrian gastronomic handicrafts. In 2000, a European Project, FINESTRA, was founded by the European Commission, aimed at improving opportunities for small and medium enterprises in peripheral areas of European countries. In that context, city logistics in Terni were analysed as possible improvements of operating condition for the several haulage local firms.
Last-mile, a procedure to set-up an optimized delivery scheme
157
The latest urban transport master plan (Municipality of Terni, 2001) then confirmed the willingness to realise a logistic base along the major transport facilities crossing Terni. This base should be located about 5 km from the railways station, having a role both of urban and regional centre. In such a positive context, within the Master Agreement between Italian Ministry of Environment and the National Agency for Energy, New Technologies and Environment (ENEA), Terni was selected as a case study for evaluating the environmental and social impacts of a re-organisation of urban goods delivering. Starting from the original idea, the Chamber of Commerce, ENEA, by mutual consent with the Ministry of Environment, decided to analyse the possible reorganisation of e-commerce deliveries, involving the Department of Transport at the University of Rome "La Sapienza" for the design. Starting from the analysis of the study area and future demand estimates, a transitpoint/drop-point scheme was adopted. Such a scheme provides e-commerce deliveries from one or more transit-points to several places spread all over the area with the following main characteristics: • public place; • easy access on foot from home; • along usual trips for many people; • enough space to stock parcels; • personnel activities and characteristics compatible with parcel picking-up and delivering. The first two characteristics assure that possible reductions in delivery vehicle distance covered would not be exceeded by the additional travel distance by private cars, which is an important feature for the environmental performance of the new delivery system. Starting from the characteristics listed above, the following public spaces have been identified to be suitable for drop-points: cafes, booksellers, post offices and newsagents. Using the AHP, the following classification was established (Table 2). Table 2 Drop Point Classifications Classification Option Total Score
1
Caft
2 3 _4
Newsagents 5,02 Bookseller 2,26 Post Office 1,83
5^9
Of course, not all the parcels are suitable to be delivered to drop-points. In the case of large volume or weight, deliveries would be "to door". The same would occur for particular customers such as handicapped or aged people. Demand estimation The total demand for BtoC good purchases for the year 2010 was forecasted by applying aggregate demand estimates predicted for Italian towns, an average of 15 purchases for each on-line shopper per year. Thus, the expected number of BtoC good purchases in Terni on 2010 is approximately 1,400 per day.
158 Logistics systems for sustainable cities The distribution of the total demand among different traffic zones has been estimated by means of a nested logit model, which reproduces the choice process in two stages. In the first stage, we determined if Internet users would buy or not. In the second stage, people that chose to buy, also chose the type of goods purchased which involves a physical delivery, and services (e.g. financial services), which does not. A specific revealed preferences survey on users' behaviour was conducted in a local Internet Cafe. A very interesting finding from the survey is that the drop-point delivery scheme seems to be very appealing (86% of on-line purchasers are willing to exchange a limited discount for not having the goods delivered to their home). Different model specifications were compared and then the form using the following variables was been selected: • age (1 if for younger than 34 years, 0 otherwise); • employment: (1 for workers, 0 otherwise); • degree: (1 for graduates, 0 otherwise); • woman: (1 for women, 0 otherwise). The following utility functions have been applied for buyers of goods, buyers of services and not-buyers, respectively: Ugoods= bix age + bi x degree + bi x employment U settees= bix woman + b$ x buy U not buying = *6 x
nobuy
Alternative-specific attributes are buy and nobuy. Calibration of the model using the maximum likelihood method based on a pilot sample of approximately 100 interviews provided estimates of the coefficients (Table 3).
coefficient b, b2 bi bi bs b6
Table 3 Coefficient Estimates Rho squared = 0.5139 estimation standard error -20.85 49.19 56.13 130.4 -8.257 31.57 853.2 134.9 -820.9 68.61 1.439 0.8915
t ratio -0.424 0.4304 -0.2616 6.325 -11.96 1.615
Results of the design procedure As inputs of the procedure, three possible locations for transit-points and a large set of droppoints have been individuated. Using the procedure described in the preceding pages, 2 transit-points and 30 drop-points were selected (Figure 5). The single zones are the influence areas of each drop-point, while the two differently shaded areas represent the influence areas of the two transit-points. The inner town is highlighted because it is a Limited Traffic Zone for low emission vehicles.
Last-mile, a procedure to set-up an optimized delivery scheme
159
Figure 5 Transit-points selection and drop-points clustering Vehicle typology selection led to two different typologies, one for deliveries out from the Limited Traffic Zone (LTZ) and another one (electric) for service within the LTZ. It is worth noting that the use of low emission vehicles in the free zone (not LTZ) was not taken into account as the aim of the analysis concerns the efficiency of the logistic scheme independently of technological improvements. The multi-criteria analysis for the selection of non-innovative vehicles was based on a comparison among six indexes: buy-out cost, annual personnel cost, daily fuel consumption and emissions of CO, PM and NOx. Rough estimations of the total distance covered, average speeds and total delivery time were carried out for each pre-selected vehicular typology in order to calculate personnel costs, fuel consumption and pollutant emissions Results of such a calculation are shown in Table 4.
Vehicle
Piaggio Ape
buy-out cost (EURO) 13,428
Table 4 Vehicle Attributes Annual Daily fuel Daily pollutants emissions personnel consumption cost (litres) (EURO) CO(g) PM(g) Nox (g) 3.2 4.9 37.1 19.5 61,675
Max
FIAT Scudo FIAT Ducato
30,987 32,020
61,975 61,675
4.9 6.5
56.2 73.9
7.4 9.7
29.5 39.0
160 Logistics systems for sustainable cities The comparison among the three typologies leads to Piaggio Ape. After drop-point clustering by means of a heuristic algorithm (GA), the better tour set for an average day of deliveries was carried out by means of a TSP algorithm so that the number of vehicles could also be defined. In conclusion, the vehicular fleet for e-commerce deliveries in Terni in 2010 should be of 2 Piaggio Ape Max and 2 Piaggio Porter elettrico. Environmental evaluation The Ministry of Environment's major aim in charging ENEA and the University of Rome with the study on Terni was to demonstrate the reduction in environmental impacts and external costs. Such a reduction has been estimated by comparing the new scenario with a possible situation without any rational measures, where all deliveries of goods purchased by BtoC ecommerce are performed by direct trips. Although the environmental benefits are quite small in absolute amount (as e-commerce deliveries represent only a small part of vehicular traffic), in percentage terms they appear very surprising: with a 55% reduction in total distance covered and more then 80% reduction in pollutant emissions! The higher percentage reduction in emissions is due to the adoption of both a more efficient delivery scheme and a more suitable vehicular typology than the current ones, that are insufficient.
FURTHER DEVELOPMENTS Further developments of the algorithm are being undertaken. One improvement concerns the structure of the design scheme, which should solve the problem of fleet sizing and composition jointly to that of transit-point location and drop-point clustering. The correspondent solution procedure should be based on a triple string genetic algorithm that also includes the optimal routing problem and the vehicle capacity into the performance function evaluation. It expected that the efficiency of the delivery scheme could be greatly improved if a more effective TSP algorithm were used. Here, the framework of the genetic algorithm might be extended by adding a further fourth string. Finally, a more general test of the procedure will be carried out on a larger size network.
REFERENCES Ambrosini, C. and G.L. Routhier (2000). Objectives, methods and results of surveys carried out in the field of urban freight transport: an international comparison. World Conference on Transportation Research, Seoul, Paper Number 2504. Ben Akiva, M. and S. Lerman (1985). Discrete Choice Analysis: Theory and application to Travel Demand, MIT Press, Cambridge, Mass., USA. Censis (1999). Un computer in ogni casa, formazione diffusa, telelavoro. www.censis.it. Dablanc, L., (1998). Freight Transport Regulation and New French Urban Mobility Master Plans. 8th WCTR Proceedings, 1, 627-636. Golob, T.F. and A.C. Regan (2001). Impacts of Information Technology on Personal Travel and Commercial Vehicle Operations: Research Challenges and Opportunities, Transportation Research, Part C, Emerging Technologies, 9, 87-121.
Last-mile, a procedure to set-up an optimized delivery scheme
161
Holmqvist, M., O. Hultkrantz, G. Stefansson and A. Wingqvist (2000). The logistical consequences of e-commerce -theoretical scenarios for spare -part distribution. World Conference on Transportation Research. Seoul, Paper Number 2202. Lin, C.H.Y., C.K. Chow and A. Chen (2002). A location-routing-loading problem for bill delivery services, Computers & Industrial Engineering- 43, 5-25. Min, H., V. Jayaraman and R. Srivastava (1998), Combined location-routing problems: A synthesis and future research directions. European Journal of Operational Research. 108, 1-15. Municipality of Terni, (2001). Piano Urbano del Traffico. Owen, S. H. and M. S. Daskin (1998). Strategic Facility location: A review. European Journal of Operational Research. Ill, 423-447. Patier-Marque, D. (2000). Which tools to improve urban logistics? World Conference on Transportation Research. Seoul, Paper Number 2508. Picard, J.M., 2000. Sfideper il commercio al dettaglio: la grande distribuzione- Proceedings of the Conference "B-distribution Forum: Gli impatti dell'e-commerce sulla distribuzione fisica", Baveno, Lago Maggiore, Italy, 11-12 September. S.T.A. (2000). Studio per la mobilita delle merci nell'area centrale di Roma. Technical report, Societa Trasporti Automobilistici, Roma, Italy. Stough, R.R. (2001). New Technologies in Logistics Management. Handbook of Logistics and Supply Chain management. A. Brewer et al. (Eds.). Elsevier Science, 513-520. Taniguchi, E., M. Noritake, T. Yamada and T. Izumitami (1999). Optimal size and location planning of public logistics terminals. Transpn. Res.-E. 35, 207-222. Visser, J., A. van Binsbergen. and T. Nemoto (1999). Urban Freight Transport Policy and Planning. City Logistics, E. Taniguchi and R.G. Thompson (eds.). Institute for City Logistics, Kyoto, Japan, 39-70. Wu, T.-H., C. Low and J. W Bai (2002). Heuristic solution to multi-depot location-routing problems. Computers & Operations Research, 29, 1393-1415.
This page is intentionally left blank
12
TOWARDS A MATCHING SYSTEM FOR THE AUCTION OF TRANSPORT ORDERS J.H.R. van Duin, Delft University of Technology, the Netherlands J.C. Kneyber, Delft University of Technology, the Netherlands
ABSTRACT This research will focus on the use of new information technology to improve urban freight transportation. Using IT in urban freight transport inter-organisationally, which can be done by facilitating the trade and negotiation between shippers and carriers, will be the focus of this research. The main goal of this research is to establish in what way matching systems contribute to the realisation of the goals of the involved actors involved. A matching system is a system that tries to combine the demand and supply of freight transport services. These 'matching systems' offer shippers and carriers the ability to communicate with each other via Internet and to trade in an innovative way. These systems are set up for various reasons, such as a better use of the available resources and a decrease of the transport costs. This research analyses the effects of these matching systems on carriers, the companies that execute the physical transportation of the freight, and shippers which are the companies that offer loads to be transported. To be able to analyse an 'artificial' auction of transport orders is modelled in Delphi. The main objects in the model are shippers, carriers and trucks. In the model a shipper offers loads. Different carriers can bid on these loads. Each carrier has a number of trucks which can transport the freight. The model is discrete dynamic. Simulation experiments have been carried out in five scenarios: The results of the simulation experiments have given us a rich dataset for analysing the auction behavioural aspects of the carriers
164 Logistics systems for sustainable cities RESEARCH SCOPE Introduction In recent years urban freight transport has faced more and more problems and challenges. The demand for transport is increasing rapidly and is expected to increase in the next decade. The road capacity cannot be expanded much more, especially within urban areas. Growing congestion levels damage the economic growth of cities. Environmental problems caused by traffic, such as air pollution, energy conservation, sound and noise hindrance, vibrations and safety problems are becoming major issues. The complexity of logistic processes continues to increase. This has several reasons for this. First of all, consumer expectations and demands are becoming higher and higher. Secondly, logistic chains are becoming more complex. Actors in the logistic chain tend to take over each other's roles. There is an increasing number of actors involved in the transportation process. The development of third party logistics has a major influence in this development. Thirdly, government policies on freight transportation are becoming more stringent, due to the growing problems this sector creates for society as a whole. A few examples of these stricter policies are the introduction of time windows for the loading and unloading of trucks and rules about the weights of trucks in the inner centres of cities. To cope with the growing problems in transportation and the growing complexity of logistics processes, many new initiatives have been proposed (Taniguchi et ah, 2001): (a) Using new information technology (b) Co-operation among carriers to increase efficiency and lower the transportation costs (c) Public terminals, such as city distribution centres (d) Underground freight transportation systems This research addresses the use of new information technology and co-operation among carriers to improve urban freight transportation. In urban freight transport, advanced information systems can be used to optimise the use of trucks and commodity flows within one company, which could be called an intra-organisational approach. Information technology can also be used inter-organisationally, which can be done by facilitating the trade and negotiation between shippers and carriers. Research shows that online auctions can achieve gross savings ranging from 5-40 percent (Tully, 2000), with an average of 15-20 percent gross savings being more typical (Cohn, 2000). Besides the cost saving, online auctions provide several other benefits, such as rapid matching demand and supply, expanding the scope of potential markets for business enterprises, improving the information flow and enhancing supply chain integration. Although security and trust issues are currently considered as challenges for online auctions (Beker, 1999 and Furnell and Karweni, 1999), more attention from industry is being paid to online auction markets because of its potential huge advantages. These 'matching systems' offer the ability for shippers and carriers to communicate with each other via the Internet and to trade in an innovative way. These systems are set up for various reasons, such as providing better use of the available resources and to reduce transport costs. In other branches matching systems already exist but on the level of city logistics little research has been undertaken regarding the effects of these systems. Therefore many questions are still unanswered, for example: (a) What kind of systems are available and how do they function? (b) What are the expectations and demands of users regarding these matching systems? (c) Do matching systems fulfil these expectations? (d) How could matching systems contribute towards satisfying customers?
Towards a matching system for the auction of transport orders
165
Research scope
This paper can be seen as a part of an ongoing research program studying the effects of using auctioning systems for the transport order management. The study was started at Kyoto University (Taniguchi et al 2002), and Kneyber (2002) added the influences of actor perceptions to the model. Since the assumptions made are very theoretical, the basis of both models will be reused in a simulation gaming environment, and later in distributed gaming environment with real players. However, during the development of the research models we want to keep track on the outcomes of the auctioning systems and therefore we will show the main results of the simulation in this paper. Still we need to realise that we have used artificial experiments and therefore the outcomes and conclusions can not be fully realistically interpreted, but will show us possible directions for adjusting the modelling components towards a real matching system. Since the most important additional development to the model developed at Kyoto University is the incorporation of actors' perspectives, we describe a detailed actor analysis of the freight transportation market. This analysis focuses on the goals and expectations of each actor regarding the matching systems, hi order to see whether matching systems can be beneficial we developed a simple theoretical simulation model in order to quantify a range of effects. This paper also describes the conceptual design of the model, the simulation scenarios and the general results.
ACTORS
Introduction There are a number of actors that are involved in the freight transportation process. The goal of this actor analysis is to describe the different stakeholders, the relations and interactions between them and their expectations and demands of a matching system. These expectations and demands are used to verify whether perceived effects of matching systems really can exist. Insights have been derived from the literature and a large interview with a Japanese company called ILNet. Since April 2001, the company ILNet has operated a freight matching system for
166 Logistics systems for sustainable cities truckloads in Japan. At this moment, shippers primarily use the system for urgent orders. Now, over 100 carriers are members of ILNet. ILNet expects that this number will increase to about 3000. Transportation is carried out for approximately 500 shippers. ILNet has permission to ship loads throughout all of Japan. The main actors are: 1. Carriers, the companies that execute the physical transportation of the freight (the truck companies). 2. Shippers, the companies that offer loads to be transported. 3. Third party logistics companies (3PL), this name covers a wide range of companies offering very different services. There are brokers, forwarders and service providers. Some of the companies can carry out the transport themselves, others just organise the transport. 4. The companies that operate a matching system. These companies can be third party logistics companies or even carriers. There is usually some kind of hierarchy between the various actors in the transport chain. It happens quite often (and more often in recent years) that shippers do not have direct contact with the executing carrier, but only have contact with a 3PL. It is also possible that this broker hires another broker to find the best freight carrier to execute the job. This makes the freight transport market a very complex market and thus difficult to grasp fully and even more difficult to model. Also the number of 3PLs is growing, which means that the area is only getting more complex.
Carriers Transport companies carry out about 70% of the professional freight transport. The remaining 30% is carried out by the industry itself. For example, in the Netherlands, there are about 12,000 road carriers. Only 75 are considered as large companies (with more than 100 trucks), 7465 companies have only 1-4 trucks. Quite a few companies have just one person. Their market share has been growing in the last few years. However, the market share of big companies, with more than 100 trucks, is also growing. This means that medium sized companies are under a lot of pressure. They cannot profit from the economies of scale, as the large companies can, and they have higher fixed costs than the small companies. Usually the small companies are hired by the large companies to execute shipments. The competition between the small companies to get these orders is fierce. The result is that transport is carried out for a very low price, since companies cannot afford to increase their transport prices even if wages or costs (of fuel for example) increase. Another problem is that there are more trucks than loads to be carried. Due to the competition on prices, many companies are enlarging their capacity to increase the amount of ton-kilometres of the company. This again increases the price competition. How to cope with the development of new distribution concepts, like Just-In-Time (JIT), has been a big challenge for carriers over the last decade. Most of the new concepts cause a more frequent delivery pattern and a lower average truckload. JIT implementation substantially affects the criteria by which carriers are selected, increases shipper-carrier communications (and communications needs), reduces the number of carriers used and leads to mode choice changes which favour truck only and truck-air transportation over truck-rail services (Leib et al, 1998). For carriers, using a matching system can have various advantages, which can improve their revenue. Firstly, they can reach new customers, because they get access to a nation-wide
Towards a matching system for the auction of transport orders
167
network of ready-to-move freight. Secondly, they can decrease their costs. This can be done in various ways: • By improving their fleet utilisation: decreasing their deadheads; increasing backhaul rates and shortening long layover • By avoiding costs and delays of credit checks for new shippers, since all the shippers that participate in the auction are supposed to be reliable and trustworthy • By reducing selling costs and reducing paperwork • By simplifying their accounts and improving cash flow
Shippers The manufacturers and distributors are continuously changing their strategies. Shippers are often operating internationally, with products being made for the global market instead of for one country. Components and raw materials are bought world-wide, from a small number of suppliers. The number of production centres and supply centres has reduced significantly. Remaining centres are therefore of a larger scale. This geographical concentration leads to larger distances. New logistics policies within companies, like JIT, are in contrast with this development. Just-in-Time requires more frequent shipping of loads, with a smaller load-sizes. Carriers are required to service more destinations and provide better services over larger distances. Many organisations concentrate on their core business activities and use specialised companies to perform their transportation and logistic activities. Carriers are asked to perform more and more different tasks, like supply control and other value-added services. This has led to the development of very different carriers, who can offer very different services. There are several expected advantages of using a matching system for shippers: (a) A better control of the product and information flow. This is realised by a better insight in the supply chain, an increase in logistics control and better visibility (b) Achievement of high performance standards (c) Reduction of transportation and transaction costs, with lower freight rates, lower loss/damage rates and lower administrative costs (d) Using the matching system might be an easier, more convenient way to move freight, so shippers can focus on their core business (e) Shippers gain a wide access to qualified carriers
Third party logistic companies This is the most diverse group of actors, when looking at their activities and services. Hence, it is difficult to give a general description of these companies and their goals. Many 3PLs specialise in only one area, such as warehousing or freight bill payment. Others handle the entire logistics processes from order-entry to final stocking of the shelf at the end of the line retailer. Crum and Allen (1991) examined the impact of logistics strategies adopted to cope with the demands of JIT systems on shipper carrier relationships. Strategies examined include carrier reduction, the use of EDI and the use of long term contracting for motor carrier services. Their study found that shippers were increasingly entering into 'partner-shipping' relationships with their carriers and that many US carriers received more than thirty percent of their revenues from a single key shipper. When they revisited the industry a few years later they noticed a continued move from transactional to contractual relationships (Crum and Allen,
168 Logistics systems for sustainable cities 1997). According to Berglund (1999) there are several indications that the 3PL industry has not reached maturity: (a) There are still a large number of 3PL providers, suggesting no clear market leaders (b) There is as yet no unique and undisputed terminology (even defining the 3PL industry itself) (c) There are few market players that concentrate exclusively on 3PL, most are subsidiaries of large transportation companies
Matching companies There are several names for matching systems; exchanges, load matching portals and electronic freight markets are only a few examples. These systems are a form of so-called electronic commerce (e-commerce) or electronic marketplaces (e-marketplaces). A definition of ecommerce is: all business activities done electronically to improve the efficiency and effectiveness of market and business processes (Dutch Ministry of Economic Affairs, 1998). An e-marketplace can be defined as an electronic exchange where firms register as sellers or buyers to communicate and conduct business over the internet (NOIE, 2001). A B2B electronic exchange is a virtual marketplace where buyers and exchange suppliers can interact and transact. Electronic markets are inter-organisational information systems that allow the participating buyers and sellers to exchange information about prices and product offerings (Bakos, 1991). Nemoto (2001) divides "business" into "shippers" (e.g. suppliers, manufacturers, wholesalers, and retailers) and "logistics service providers"(e.g. freight carriers, warehouse firms and third party logistics). B2B usually means the transactions between shippers (S2S) and distinguishes the transactions in the market of logistics services, between Shipper and Logistics service provider (S2L) and between Logistics service-providers (L2L). In this paper we define companies that operate such matching systems as matching companies. Such a matching company tries to match the capacity of carriers at a certain time in a certain region with available loads that need to be transported at that time in that region. The idea of electronic matching systems has only been developed recently and the market for these systems is changing and developing constantly. According to NOIE (2001), firms now tend to enter into arrangements that encompass existing business relationships, such as "private e-marketplaces", smaller electronic exchanges revolving around existing business relationships within their sector. The value proposition of the private e-marketplace is to provide customers with buy and sell services to enhance their current way of doing business, grow their efficiency and therefore staying competitive. This involves companies taking their existing processes and trading networks online to gain the connectivity and speed of the internet, within a secure environment, thus guaranteeing that sensitive information and processes are not exposed to the world. Although a matching system could offer significant advantages, these kind of systems are less common practice than might be expected. There are three possible reasons for a company not using a matching system: 1. Unawareness such systems exist 2. Uncertainty about working with an unfamiliar system 3. Doubts about the benefits of the system First of all, some (small) companies are possibly not familiar with the existence of these systems or maybe think they are something for the distant future. Companies can be afraid to
Towards a matching system for the auction of transport orders
169
use these systems for several reasons. Carriers might question the number of loads that will be available for them to transport and shippers may doubt the number of carriers that will bid for their loads. The companies can be hesitant to work with unknown companies, since the carriers and shippers they work with now are trustworthy and reliable. Carriers are not guaranteed that they will get their money from unknown shippers, shippers might question whether the carriers will actually transport the load safely and on time. Moreover there are costs for using the matching system. Nowadays most of the existing systems can be used for free or for a small amount of money. Companies cannot be sure that this will still be the case in the future, especially now the bubble of the internet economy seems to have burst. Carriers and shippers do not want to be dependent of companies or systems whose behaviour is not very predictable. A company could suddenly raise the price of using the system, or it could go bankrupt, as quite a few companies that operate matching systems have done over the last few years. Carriers and shippers do not like the idea of sharing their information. They feel that information on their trucks, loads and prices is confidential and that using a matching system might give their competitors insight into their operations, giving those competitors an advantage. Companies might also think that the matching system couldn't give them enough advantages. With the fierce competition in the transport sector, carriers have to cope with over-capacity. Therefore, shippers can easily find carriers with free transport capacity without having to resort to a matching system. Many companies think that improving the internal planning and their relations and communication with customers is more beneficial than searching for new and unknown customers, even though this may be cheaper. Until now it has been difficult for matching companies to get customers, because the eventual advantages of using these companies are unknown. Carriers and shippers mostly choose the traditional way of transporting goods, because this has proven to be reliable and workable. At the moment, planning is done by professional and experienced people, who know the transport world and have good relations with their customers. It may very well be possible that they can get results as good as a matching system can.
Summary of the perceptions on matching systems Currently, relations between shippers and carriers are rather fixed. Shippers tend to use the same carriers for their freight and carriers have contracts with fixed shippers to execute their transport. Quite often there are long-term contracts between these parties, in which is agreed how much freight will be transported per day/week. It is rare for a shipper to extensively source the market in search for a cheaper carrier for the load they need to get transported. With the arrival of third party logistics shippers usually contact a carrier or 3PL to ask whether they can transport the freight or search an appropriate carrier for them. Such a carrier or 3PL makes sure that the load will be shipped, either by itself or by another carrier. General benefits of a matching system are: (a) The digital marketplace will improve market efficiency to the benefit of buyers and sellers, leading to more market transparency and a reduction of arbitrage and price differentials between markets (b) Commercial intermediaries who do not add value to the market, but have only survived in the past because of market inefficiency, will have to adapt or perish (c) Digital marketplaces will replace or enhance existing distribution channels, replacing those traditional intermediaries who choose not to adapt
170 Logistics systems for sustainable cities (d) Digital marketplaces also offer the industry the opportunity to improve transaction efficiency and cut embedded sales, purchasing and administrative costs (e) Digital marketplaces provide buyers and sellers with a more efficient vehicle to identify new business opportunities General perceived risks and disadvantages of a matching system: (a) Shippers and carriers are unfamiliar with such systems (b) Shippers and carriers are afraid to use these systems, because: (i) They do not trust the company and the unknown users of the system (ii) They do not want to depend on a possibly unstable and/or unreliable company (iii) Shippers and carriers might think that current practice is better than such a system This section raises two issues that are important for building a model of a matching system. First, the model should provide insights as to whether the advantages of matching systems are significant. The outcomes of the model should focus on efficiency and costs, with regard to carriers, shippers and the system as a whole. Second, it is likely that companies, when starting to use a matching system, will only use it for part of their loads, partly because of obligations towards regular customers and partly because they are reluctant to depend completely on a matching company for their income. The model should encompass both orders assigned by a matching system as well as orders acquired in the traditional way.
MODEL DESCRIPTION
Existing models Taniguchi et al, (2002) researched the effects of a matching system on traffic flows. In this model a situation with a matching system is compared with a situation without a matching system. Three carriers are modelled, each with four trucks. The infrastructure is made of a grid-network of 25 nodes. Every carrier has one of the nodes as the location of their DC (distribution centre), which is the starting and ending point for the routes of the trucks. Every carrier is assigned a (random) number of customers. This means that randomly a number of nodes is selected and that a time window is attached to these nodes. These so-called fixed customers represent the customers that the carrier has found without the matching system. Then, the carrier makes a transport schedule for these fixed customers, a plan for the trucks to visit the various nodes, preferably within the time window of each node. At every customer, the truck has to wait a short time, which represents the loading and unloading. The schedule of every truck begins and ends at the home base of the carrier. The model uses costs for the driven distance (fuel costs), costs for the travel time (costs of the driver) and fixed truck costs. There is also an important role for penalty costs. If a truck arrives at the customer before or after the time window, then the truck incurs a penalty. Arriving too late results in a higher penalty than arriving too early. After making a transport schedule for the fixed customers, a 'new customer' is defined (randomly choosing a node and a time window). Every carrier then makes a new transport schedule, for their fixed customers and the new customer. The new customer is assigned to the carrier whose total distribution costs for the new schedule are the lowest. The new customer is assigned to the carrier with the smallest extra costs for executing the job of the extra customer.
Towards a matching system for the auction of transport orders 171 In other words, the carrier with the smallest difference between the total costs of the old schedule (fixed customers only) and the total costs of the new schedule (fixed customers and the new customer). The carrier with the lowest penalty costs gets to execute the transport of the new customer (if the penalty costs for two or three carriers are equal, the criterion of scenario 2 is used). With help of the Tabu-search method for several situations (locations and time windows of fixed and new customers and locations of carriers) the optimal schedule is determined and the load is awarded to the 'best' company. Differences in models In the model developed at Kyoto University, time is not simulated. In every simulation, only one new load is modelled. Instead of trying to fit in various loads in the schedules of the trucks, the simulation of scheduling one load is repeated many times (with different fixed customers and different new loads). No attention is paid to the negotiations between the shipper and the carrier, the carrier with the lowest costs automatically 'wins' the load. In Kneyber's (2002) model, the negotiations take place as in a real auction, while each carrier can make a bid The shipper chooses the best bid, after which the carriers can decide to make a new bid. Other factors that the carrier can use to decide what his demand price will be, like the profit they want to make, the number of transport orders they have already won and the preference for transport within a certain region, are not considered in the model developed at Kyoto University. Also other criteria the shipper might use, like pick up time, transport speed, punctuality of the carrier and reliability of the carrier, are neglected in the model. Kneyber's model uses two distinct factors to define a bid: not only the price, but also the difference between the possible pickup time (carrier) and the preferred pickup time (shipper) is considered. Other factors are also used in determining the bid price. Another difference between the two models is in the planning of a new schedule: in the model developed at Kyoto University, adding the new load to the schedule means making a whole new schedule for all the loads and all the trucks. In this research, only a new schedule is made for the truck that will transport the load. Finally, in Kneyber's model the focus is less on the performance of the total system and more on the behaviour and the performance of individual carriers and shippers.
Conceptual description of the simulation model
Figure 2 Information flows between the actors and the matching system In Figure 2 the relations between these actors are described more specifically. For one load the following information flows can be distinguished in chronological order:
172 Logistics systems for sustainable cities (1) (2) (3) (4) (5) (6)
A shipper announces they have a load for the auction The carriers receive information about this load The carriers send their bids on this load to the matching system The shipper receives these bids and chooses the best bid The shipper announces which bid in his opinion is the best This information is sent to the carriers
The carriers then can make a new (better) bid on this load if they want to. These new bids then are sent to the shipper, who again evaluates them. This cycle of processing steps 3 to 6 continues, either until no carrier makes a new bid or until a certain amount of time has gone by. Then the auction ends and the carrier who at that time has made the best bid transports the load. Carriers acquire only a part of their transport orders via the matching system. Each carrier should already have a number of orders at the beginning of the simulation. These transport orders which are not obtained via the matching system, are called fixed transport orders. The carriers receive all their fixed orders at the beginning of the simulation. How the carriers acquire these fixed orders and how they negotiate about them is not a part of the simulation. When the simulation begins, each carrier makes a schedule for their fixed orders. This schedule describes how the truck will execute the transport of the loads. In the schedule four different activities are defined: loading, unloading, driving and waiting. The loading takes place at the pick up point of the load. Loading and unloading times grow linearly with the amount of freight that has to be transported (and therefore loaded or unloaded). Unloading takes place at the distribution centre of the carrier. Driving is the time used to get from one (loading or unloading) location to another loading or unloading) location. At such a location freight might be loaded onto the truck or the freight of the truck might be unloaded. If a truck is not busy with one of these three activities, a truck is said to be waiting. This means that the truck is not doing anything until the next activity. At certain (random) times, a shipper announces to each carrier that they have a load to be transported. This means that they have some freight that they want one of the carriers to pick up and bring to their distribution centre. They ask the carriers to bid on this load in an auction. The shipper gives information about the load to the carriers, then announces how large the freight is (tons), the pick up location, the preferred pick up time and at what time the auction will end. This announcement is called a "load announcement". When a carrier receives a load announcement, it must decide whether it will bid on this transport order, and if so how high the bid will be. First, it tries to schedule the load at one of the trucks. The carrier searches for the truck which can pick up the load for the least extra costs, but it also looks at the pick up time which is associated with those extra costs. Every carrier has a trade off function for costs versus pick up time. Some carriers will choose the truck that can pick up the load for the lowest costs and do not look at the pick up time, other carriers only choose a cheaper truck if this does not have much effect on the pick up time. After the carrier has decided what its 'minimum' costs are, and what the associated pick up time is to transport the load (note: as mentioned, these 'minimum' costs are not necessarily the real minimum, they are the minimum of the costs adjusted for the difference with the preferred pick up time by using the trade off functions characteristic for this carrier), the carrier will decide how to bid for this load. The carrier can use the following factors to determine its bid: (a) Time-Cost function (the order response time and the related cost-price calculations), planning method (b) Income already won (can influence the responsiveness in providing discounts, Table 2) (c) Loads already won
Towards a matching system for the auction of transport orders
173
(d) Preference to have a full schedule ASAP and number of bids done on this load (can influence the profit-bonus, Table 2) (e) Profit Table 1 Example responsiveness-table Income Time until transport ill already won < Vi day day -15% <1000 -10% 1000-2000 -10% -7% 2000-3000 0 0 3000-5000 +2% +0% >5000 +5% +2%
Table 3.2 Example profit-bonus 1 1-2 days -8% -5% -2% 0 0
>2 days -6% -3% -2% -1% 0
# bid on this load 1 2 3 >3
profit-bonus 10% 8% 7% 5%
Performance indicators Several performance indicators are defined to analyse and compare the results of the simulations. These performance indicators will be discussed per actor. For the carrier, the number of orders won is of course very important since it is strongly related with profit (Table 2). Another characteristic is the number of bids the carrier makes. This indicator can be specified into the number of times the carrier bids one time on a load, the number of times they bid twice on a load and the number of times they make three bids for the same load. These indicators can show how often a carrier makes a bid on the same load and has also a direct link to the profit gained by the application of the profit-bonus table. Also the characteristics of the bids are analysed: what is the average bid price and what is the average difPUT of a bid? difPUT is the difference between the preferred pick up time of the shipper and the pick up time the carrier offers in his bid. For the orders won the average price and the average difPUT are analysed also. For the trucks, first the total time used per carrier on the different activities (driving, waiting, loading and unloading) is calculated. For the shipper the costs of transportation are calculated. Both the total costs and the average costs per order are calculated in the model. Also the total difPUT of these orders and the average difPUT per executed load are analysed. Finally, the total number of loads the shipper announces and the percentage actually auctioned orders are calculated. Also, for society as a whole some data is used, namely the amount of time used in the simulation per activity and the total number of kilometres trucks travel to execute the transport.
Modelling assumptions 1. Value added services which carriers or 3PLs can offer the shippers are not simulated in the model. The model focuses on the physical transport of the load from A to B and the conditions under which the carrier executes the transport. 2. 3PLs are not considered in the model. Only companies that actually can execute the transport with its trucks are considered as carriers. 3. The model describes only the freight for which the shippers use external carriers. In this simulation there is no freight that is transported by the shipper himself. 4. The freight itself is not simulated. Only the negotiations for the loads and the driving of the trucks are modelled.
174
Logistics systems for sustainable
cities
5. No attention is paid to traffic flow, congestion,., etc. Driving times from A to B are static and equal for every carrier. 6. The shippers and carriers do not have direct contact with each other, all information goes to or comes from the matching system. There is no communication between the matching system and the trucks of the carriers. All communication with the trucks is done by the carrier.
SIMULATION EXPERIMENTS
Scenarios in the simulations •
Scenario 0 The first scenario can be interpreted as a reference case. In this case we have specified the three carriers with the same characteristics. They all have identical cost-time functions and they apply the same responsiveness table (The responsiveness is dependent on already won income and the time between transport order and real transport). The simulation has about 600 transport orders and about 350 orders have been accepted by the auction. The average number of bids was about 1100. The total incomes for the three carriers have been analysed.
•
Scenario Random Assignment In this scenario no bidding activities are carried out by the carriers but the transport orders are assigned to the first carrier who can handle the order.
•
Scenario 1. Carrier behaviour —price & time accuracy The bid price and offered timing accuracy (difPUT = difference with PickUpTime) are changed in this scenario for each carrier. The hypothesis to be tested is that the scenarios in which the price and/or the difPUT of carriers' bids are increased, less orders will be won by these carriers. The goal of the simulation is to investigate in what way a price increase and a difPUT increase influence the number of orders a carrier wins. Not only will the absolute effect of price and difPUT be investigated, but also the effects of a price increase will be compared with the effects of a difPUT increase.
•
Scenario 2. Carrier behaviour - responsiveness In scenario 2 the responsiveness of the carriers is changed. The responsiveness has a direct influence on the price of a bid via the so-called responsiveness-table. Four different attitudes are defined: (1) The carrier is always equally nervous. (2) The responsiveness increases when the time until transport is shorter. (3) The responsiveness decreases when the income already won is higher. (4) A combination of option 2 and option 3, when the time until transport is shorter, the responsiveness increases, when the income already won is low, the responsiveness is higher. In this option the used values are equal to those used in scenario 0.
• Scenario 3. Carrier behaviour - Cost-time-functions In scenario 3 the slopes of the cost-time functions of the different carriers are changed. This means that some carriers think that price is more important than punctuality and some carriers think that punctuality is more important than price.
Towards a matching system for the auction of transport orders
175
Summary results The general results from the scenarios are summarised here for the different perspectives in order to judge whether expectations and possible advantages really could be met. The perspectives are workout for shippers, carriers and society as a whole. Carriers Carriers try to maximise their profit. They try to achieve this by maximising their turnover and minimising their costs. Turnover can be increased by a high price per order or by winning a larger number of orders, for which the transport of the carrier needs to meet the demands of the shippers. The costs can be decreased by a fuller schedule, a lower number of vehicle kilometres and by lower search and negotiation costs. The simulation shows that average bid price is lower than the average expected price and that the average price per auctioned order is lower than average bid price. This indicates that the auction has a price decreasing effect and understates the price decreasing effects mentioned by Tully (2000) and Cohn (2000). The average difPUT per auctioned order is lower than the average difPUT per bid, which indicates that the auction has also a decreasing effect on the difPUT. A carrier wins less orders if its price is higher. But because such a carrier still wins a significant number of orders, the income of this carrier increases. When maximisation of the income/profit is the main goal, a strategy with a very high price seems to be favourable for the carrier in the simulations. From this the conclusion can be drawn that a low transport price only can be realised if enough carriers participate in the auction and if the participating carriers have significant free space in their schedules. When few carriers participate, there is not enough competition to speak of a real market price. A higher difPUT means a lower income than the competitors, but it does mean a higher income than with a lower difPUT would be the case. The simulation shows that a higher difPUT also means that the average price increases. The negative effects of a high difPUT are relatively smaller than negative effects of a high price. A higher price or higher difPUT almost always results in more driving-time. A longer time between the announcement of a load and the preferred pick up time of this load results in more bids and more auctioned orders. If the responsiveness depends on the income already won then the carrier bids more often twice or thrice on the same load. When the responsiveness depends on the time until transport, the carrier needs the most time for driving, loading and unloading. When the responsiveness depends on both the time until transport and the income already won, the carrier needs the least time for driving and unloading. When responsiveness is always high (independent of both the time until transport and the income already won) the reduction in waiting-time is the greatest. When the responsiveness depends on the income already won, the carrier needs the least waiting-time. A higher priority for difPUT causes a carrier to win more orders. A high priority for costs causes a carrier to win fewer orders. This can be explained by the greater competition between the carriers on price than on difPUT. Shippers As already mentioned the simulation shows that the auction causes the price and the difPUT to decrease. Only half the orders available on the auction are actually auctioned. This could be a warning for shippers, not to be too dependent of the matching system. The shipper cannot be sure that their load will be transported by a carrier found via the matching system, so they should be able to find a carrier in a different way.
176 Logistics systems for sustainable cities In the model, a shipper is not able to refuse a bid made on their load. A shipper has to choose the carrier that makes the best bid, even if this bid means a very high price and/or difPUT. As a result, the shipper sometimes has to pay a high price for the transportation of their load, which of course is not in the interest of the shipper. To prevent this, before the auction begins it has to be agreed that the shipper can also decide not to accept any of the bids made by the carriers. A longer time between the announcement of a load and its preferred pick up time causes the number of auctioned orders to rise significantly. The conclusion can be drawn that this period has to be long enough for the carriers to schedule the load, to make several bids and to complete the auction correctly. Less competition on price and/or difPUT results in a longer time necessary per order for driving, loading and unloading together. Concluding in every scenario in which carriers have a higher price and/or difPUT means more driving-time, less waiting-time, more loading time and more unloading time is needed than in scenario 0. Society Taniguchi et al. (2001) defined city logistics as the process for totally optimising the logistics and transport activities by private companies in urban areas while considering the traffic environment, the traffic congestion and energy consumption within the framework of a market economy. From this, a goal of society concerning freight transport is assumed to be to reduce driving time and the number of kilometres driven by trucks. Another goal of society is that there is a fair market for the shippers and carriers. The market needs to be open and transparent for all parties involved. As already stated, a higher price almost always means more driving-time and loading-time. A higher difPUT always means more driving-time and almost always means more loading-time and less waiting-time. In total in every scenario in which carriers have a higher price and/or difPUT means more driving-time, less waiting-time, more loading time and more unloading time is needed than in scenario 0. The two carriers with the lowest priority for difPUT need more time per order for the activities driving, loading and unloading. The carrier with the highest priority for difPUT needs less time for these activities. The cause of this is unknown. The extra focus on time probably makes the scheduling more efficient by attracting more orders and therefore more efficient clustering is possible.
CONCLUSIONS The question whether a matching system eventually has positive consequences, is difficult to answer unambiguously. The general rule is that more market competition is a positive thing. The competition on the market is improved with a matching system because it helps to create a more open and transparent market, with shippers having more information on (potential) carriers and the other way around. In our research the inclusion of more realistic criteria for shippers and carriers has allowed a richer understanding of the effects of auction systems for city logistics. Therefore, we will continue to extend our research towards more realistic actors by developing a gaming environment based on agent-based modelling for the auctioning and planning of transport orders to learn more concerning the effects of these electronic marketplaces on the one hand, and on the other hand to serve as a training tool to overcome the major hesitations that shippers and carriers have in using these systems.
Towards a matching system for the auction of transport orders
177
REFERENCES Baker, C.R. (1999). An analysis of fraud on the Internet, Internet Research: Electronic Networking Applications and Policy, 9(5), 348-60. Bakos, J.A., (1991). A Strategic Analysis of Electronic Marketplaces, MIS Quarterly, 15, 3, 295-310. Berglund, M.P. van Laarhoven, G. Sharman and S. Wandel, (1999). Third-party logistics: Is there a future? International Journal of Logistics Management, 10(1), 59-70. Cohn, L., D. Brady and D. Welch (2000), B2B: The Hottest Net Bet Yet?, Business Week, January 17, 36-37. Crum & Allen, (1991). The changing nature of the motor carrier-shipper relationship, impacts for the trucking industry. Transportation Journal, 31(2), 41-54. Crum & Allen, (1997). A longitudinal assessment of motor carrier-shipper relationship trend, 1990 vs. 1996. Transportation Journal, 37(1), 5-17. Dutch Ministry of Economic Affairs, (1998). Electronic commerce action plan, The Hague, http://www.minez.nl/publicaties/pdfs/05R38A.pdf Furnell, S.M. and Karweni, T., (1999), Security implications of electronic commerce: a survey of consumers and businesses, Internet Research, 9(5), 372-382. Kneijber J.C., (2002). Modelling the matching: The Simulation of a Matching System for Shippers and Carriers, Master thesis Delft University of Technology, 128pp. Leib, R.C. and R.A. Miller, (1988). JIT and corporate transportation requirements. Transportation Journal, 27, 5-10. NOIE (National Office for the Information Economy) (2001), B2B e-commerce, Australian Ministry for Communications, Information Technology and the Arts, Canberra. Nemoto, F., J. Visser and R. Yoshimoto, (2001), Impacts of information and communication technology on Urban Logistics System, Faculty of Commerce and Management, Hitotusbashi University. Taniguchi, E., R.G. Thompson, T. Yamada and J.H.R. van Duin, (2001). City Logistics Network Modelling and Intelligent Transport Systems, Elsevier, Oxford. Taniguchi, E., T. Yamada and Y. Naka, (2002). Modelling effects of logistical matching systems on transport, Urban Transport and the Environment in the 21st Century VIII, WIT Press, 407-416. Tully S., (2000). The B2B tool that really is changing the world, Fortune 141(6), 20 March 2000.
This page is intentionally left blank
13
SYSTEMS THEORY, COMPLEXITY AND SUPPLY ORGANIZATIONAL MODELS T O ENRICH CITY LOGISTICS: AN APPROACH
Jose Mexia Crespo de Carvalho, ISCTE — University of Lisbon, Portugal
ABSTRACT This paper establishes a connection between systems theory and logistics presenting a city logistics model in order to enrich the available frameworks to study city systems. The proposed model uses two wide variables: relative entropy when applied to city systems and the number of food selling points that can be observed in the same city system. With this model one can easily classify city systems, obtaining four major categories: modern-dispersed, modemincomplete, traditional-complete and mixed city systems.
INTRODUCTION The main objective of this paper is to establish a connection between systems theory and the study of a city system using logistic related variables. The connection should be developed, essentially, at the conceptual level, trying to create a model in which the variables make the representation of a city system and, consequently, a way of thinking in terms of logistical solutions. Under these conditions, two core aspects are fundamental to structure of the whole proposal: 1) the city should be seen as a set of human, physical and informational elements in permanent interaction between themselves and the exterior, therefore as a dynamic and open system 2) systems theory does not allow but global representations of an identified or discovered set of elements, being more adequate for an ideographic research.
180
Logistics systems for sustainable cities
Therefore, systems theory criticises the basic assumptions of positivism, relating to the assumed ability to understand partial processes. Instead, the core idea of systems theory applied to city logistics should be that cities, as organisations, must be studied as a whole, since the different parts of the system interact and can, therefore, not be separated from each other. Systems theory attempts, then, to formally describe the logic of a system, i.e., its own organisation: in this case a city system. In order to do this, two aspects have been explored in this paper, on one side, systems theory in a condensed and sufficiently exemplified form, while on the other side, variables that, being part of any city system, may serve as representatives when it is intended to explain their physical, informational and human flows. Bottom line variables that are able to develop a model of basic elements have been used for analysis and explanation of a city system. Being a conceptual work, the proposed model should be, afterwards, applied to specific cities, so that it can be validated and enriched. At this stage of the investigation, only hypothetical examples have been used but in the near future, they will be empirically validated. Contributing to alternative ways of thinking has been considered fundamental, when developing this paper, in order to explore city systems under logistics principles.
INSIGHTS ABOUT SYSTEMS THEORY AND CITY SYSTEMS Fundamentals of Systems Theory were developed by Forrester (1959, 1961 and 1989) and initially called industrial dynamics. One of the major impacts of Forrester's work was the recognition that only system approaches could provide a deep insight on complex dynamic behaviours such as the bullwhip effect (Lau et ah, 1997). The network configuration of today's supply chains with an ever-increasing number of partners, each one following its own strategy and taking actions that do not always fit with the common objectives, the fuzzy nature of many of their inter-relationships and the dynamic character of all the variables are imposing a system approach too, specially if one looks for an effective supply (demand) chain management (SCM) (Simchi-Levi et al., 2000 and Sterman, 2000). As a result, the modern concept of a system, formulated around 1930 by von Bertalanffy— following previous works by Leduc and Bogdanov (Bertalanffy 1968), spread out to all fields of knowledge and became the current scientific paradigm. Among the main features that most systems share is important to emphasise (Le Moigne 1977): (1) every system has a structure, which is the set of its components plus the set of rules defining their interrelationships; (2) every system has a teleological character, i.e., is driven towards a set of goals, which is equivalent to say that every system has a purpose; (3) every system interacts with its environment, performing activities that change the system's state in order to approach the established goals. Systems are, then, intricate entities that exhibit detail and dynamic complexity. Thus, assuming both the holistic character of the system and its complex nature it is somehow usual to recommend and apply systems theory to several situations (Checkland et al. 1999, Gharajedaghi, 1999 and Sterman, 2000). City systems, considered as a whole, may, certainly, be included in one of these mentioned situations.
Systems theory, complexity and supply organizational models
181
Distribution strategies, network configuration, organisation design, supply chain integration, strategic alliances and information systems design are some of the strategic decision making processes where a powerful set of methodologies and techniques (Christopher, 1992; Gattorna et ah, 1996, Simchi-Levi et ah, 2000) using scenario planning, simulation and, generally, systems theory can offer a valuable contribution in many circumstances and situations. Many other approaches and applications are reported in the literature, namely at the entrepreneurial level and covering matters like the supply chain dynamics (Swaminathan et ah 1998, Fukunaga et ah, 2000, Groothedde, 2000 and Anderson et ah, 2000), partners' management (Sobrero et ah, 2002), performance analysis (Cakravastia et ah, 1999), decision policies (Strohhecker, 2000) and the integration of product management with order strategies (Barlasefa/., 1996). Having this in mind one should stress the importance of the systems representation of in order to study it. The concept of a model allows a simplified representation of a system and it is very useful to explain a context and their multiple problems (a city system, for instance) (Forrester 1961; Sterman, 1991 and 2000). Developing a model is thus an evolutionary complex social process that helps the group of modellers: (1) (2) (3) (4) (5)
to be explicit about their ideas about the context and its problems, to identify the features and variables that are relevant to the problem, to achieve a consensus about the problem's nature and delimitation, to gain an insight about the system's structure, and to build a construction accepted as a valid representation of the real system at a given level (Simon, 1981 and 1990).
In our times of globalisation of economies, markets and knowledge, the boundaries of the firm are fuzzy, the extended enterprise appears as the new industrial paradigm and supply chain management is recognised, if not yet logistics network management (Carvalho, 2001) as an important piece of corporate strategy (Gattorna et al. 1996). A network is a complex system of entities and relationships, processes and flows, and therefore cannot be managed—either at a strategic or at an operational level, without relying upon a system approach. Perhaps one has, today, all the means to quickly overcome distances, either in the digital or in the physical world. However, traffic growth has become a barrier in both worlds, in processing data but, more importantly, exchanging physical flows. Cities, for instance, are complex systems where all kind of physical barriers have been built and human restrictions and implications occur at any time, imposing unnatural developments and disruptions both in physical and in informational flows. As a result, one may think that cities are interesting arenas to explore some type of supplydemand organisations, having in mind central and peripheral infrastructures, overnight flow connections, bulk cargoes delivered to multi-purpose warehouses in the periphery, crossdocking opportunities within this infrastructure and transport approaches among others, using both large vehicles to concentrate flows and small vehicles to perform fast and collaborative deliveries. City constraints and logistics theory may open enormous opportunities to explore some type of supply-demand organisations within cities, envisaging collaborative models between private
182
Logistics systems for sustainable cities
companies and public municipality organisms. This type of collaboration may be the basis on which to establish relational networks able to induce flexibility and responsiveness ability into complex city systems. For instance, a strong urban centre, almost unique in commercial terms, and multiple urban centres with commercial and human integration may represent the way cities are structured and, consequently, should pilot some prospective answers that can be expected in terms of the whole urban system's supply (or demand) network. The idea will always be to optimise flows and integrate physical and information components in a city system, trying to create a more human centric arrangement. Therefore, in order to do this one should try to explore the structure of the city system, its teleological character and the interaction with the environment. The modelling approach can be, under these circumstances, a tough enabler. Thus, a systems perspective and its combination with logistics reasoning may result in a strong approach to analyse, comprehend and, in the future, to optimise complex urban economies. Having this in mind, a simple contribution approach, a prospective city logistics model, faced as an enabler, is introduced and described in the second part of this paper.
A CITY LOGISTICS MODEL One could admit several variables to structure the intended city logistics model but, once the model reaches a certain level of stabilisation and the acceptance by the academic community, it is important to search for subsequent logistical related scenarios and prospective holistic approaches and solutions. Simple variables could be used, like transportation kilometres, a conventional parameter used in urban analysis. Nevertheless, this variable, like many others, does not seem to reveal the real structure of the system, its theological character or the relations with the environment. The modelled design of the city - in order to develop a global logistic approach, cannot be obtained throughout simple and isolated variables. An effort to propose a combination of simple but measurable and useful variables to represent the system and its characteristics was the process pursued. Simultaneously, global variables with graphic representation were preferred when directly compared with others without these characteristics. Thus, variables should be global in the sense that they can best possibly explain the city system in terms of supply, geography and logistic typicality and, also, contributing to exploring the structure of the system, its teleological character and interactions with the environment. A large group of logistic related activities were systematically studied and, also, its countless variables, in the sense that two variables could be conjugated allowing a representation with the desired characteristics, such as simple global character, and good and visual illustration. Those large groups of logistic related activities allowed the study of several different variables even though of a similar nature, i.e., with the faculty of explaining, somehow, the physical and information flows structure emerging in an organisation, such as a city: (1) transportation, (2) storage and selling points, goods processing and types of movements executed, (3) handling and types and requisites of handling along the physical flows,
Systems theory, complexity and supply organizational models
183
(4) location and management of the basic infrastructure for the physical flows processing, including warehouse activities of inbound and outbound goods, and (5) logistic information systems activities. By election, and also by exclusion, one should arrive at the conclusion that the two variables that best explain urban complexity and the structure of its supplies are: (1) the number of selling points of food nature per 1000 inhabitants (PSN' 1000), and (2) the relative entropy (RE) of a city system that allows to measure the space spanning of its selling points. Both variables are strongly related to the logistic activity groups and with enormous physical temperament, thus being simultaneously global, with graphic representation possibilities and contributing to explain city systems characteristics. In this way one may say that the nature of the physical flows is even more critical then the information flows, both in the explanation of a city system and in the search of a logistic related approach/solution. The first variable, the number of selling points of food nature per 1000 inhabitants, can translate well the evolution and the complexity of the commercial urban tissue, more or less integrated, more or less developed, more or less atomised, among others. The second variable, the relative entropy of a city's system allows, in a certain way, to measure the space spanning its selling points. In fact, the entropy of a city system can be measured having as reference the formula used to measure the loss of information in a message, or even so, the disorder of the elements of a system in a thermodynamic process. The number of selling points of food nature per 1,000 inhabitants (PSN'1000), referred, year after year, by EUROSTAT and by ACNielsen, among others, points to an actual average value, in Western Europe, around 0.8 selling points per 1,000 inhabitants. Due to the concentration of selling points in urban systems, regions and countries, namely in those with less developed commercial tissues, the average value in terms of graphic representation should be 1.0 selling point per 1000 inhabitants, with the theoretical minimum of 0.0 and the maximum equal to the maximum value obtained by effective calculus for a certain city. Concerning the concept of entropy, this was extracted from the theory of information (Shanon and Weaver, 1954), since their work allowed the definition of entropy as an uncertain measure of the nature of a particular message, namely in function with the previous message. In this context, entropy becomes maximum when the uncertainty is stronger and every possibility has the same probability of happening. If in this paper we substitute the probabilities, p, by frequencies, / , entropy can become a measure of space dispersion. The entropy formula based on frequencies will be represented by the following expression:
(1)
With: E = entropy; K = number of geographic zones of commercial nature for a certain city;
184 Logistics systems for sustainable cities ft = the frequency of selling points in a certain geographic zone, knowing that (f, = w/N). Hi = number of commercial selling points in Zone i; N = total number of selling points; and The relative entropy allows one to obtain a simple measure for analysis and possible comparisons, therefore making possible a better way to explain the city system being studied. Furthermore, it allows a representation in an interval [0, 1], being the maximum entropy equal to 1,0 and the intermediate value easily identified as 0.5. It is considered that the relative entropy (RE) is represented by the following expression: RE = E/log k
(2)
When wanting a graphic representation of the two variables, so that they can explain a certain city system and its associated hypothetical logistic organisation, one can use a simple matrix of double entry, as shown in Figure 1.
Food Selling Points Number per 1000 inhabitants (PSN'1000) Figure 1 City System Evaluation Matrix From this matrix one may consider four quadrants, A, B, C and D which are explained below. Quadrant A represents strong relative entropy and a low number of selling points per 1000 inhabitants, suggesting that the city system comprehends a good selling points' geographical distribution and it presents a low coverage of selling points per 1000 inhabitants. This is a dispersed and little concentrated urban structure, with the presence of several urban poles or multiple centres. It shows recent geographic commercial planning; therefore, it will be designated as MODERN-DISPERSE city. As an example, Vigo, in the North of Spain, could be considered a city of this type. Quadrant B represents weak relative entropy due to a low number of food selling points per 1000 inhabitants, suggesting that (1) the city system comprehends low geographic distribution of selling points and (2) presents a low coverage of food selling points per 1000 inhabitants. It's a concentrated urban structure where polarisation assumes some relevance; this may originate big commercial spaces in more or less delimited zones. It shows a recent commercial
Systems theory, complexity and supply organizational models
185
geographic planning but still a weak territorial coverage, therefore designated as MODERNINCOMPLETE city. As an example, Brasilia, in Brazil, could be considered a city of this type. Quadrant C represents strong relative entropy due to an important number of food selling points per 1000 inhabitants suggesting that (1) the city system comprehends a good geographic distribution of selling points and (2) it presents an elevated coverage of food selling points per 1000 inhabitants. It is an urban structure typically integrated without large polarisation effects. The relevance here goes to the small commercial spaces in dispersed zones; one may presume an antique commercial tissue and the presence of a complex urban net. It denotes an old planning and a very complex and traditional commercial geography; in this context the city should be designated as TRADITIONAL-COMPLETE. Lisbon, in Portugal, serves as an example of this type of cities. Quadrant D presents low entropy due to a strong concentration of food selling points per 1000 inhabitants suggesting that (1) the city system comprehends a weak geographic distribution of selling points and (2) presents a high density of selling points per 1000 inhabitants. It's an urban structure that grows at several different speeds with different thoughts and times. In the most polarised zones small and antique commercial tissues are still present, denoting both zones of complex urban nets and zones of less complexity, with a more recent commercial geography. It shows an old planning and a commercial geography of great complexity and tradition, although some of the city's system areas already present emerging renovated structures. This city should be designated as MIXED city. Faro, also in Portugal, serves as an example of this type of city.
LOGISTIC CITY PROSPECTIVE: SUPPLY APPROACHES/SOLUTIONS Considering the matrix in Figure 1 and the possibility of previewing different logistic designs to supply several city systems, one should consider that cities live according to their own selling points organisation. This means that commercial structures have the option to sign, or not, contracts with external entities that render logistics services, either of private law (current operators with private investment) or of public nature (by the public administration's initiative or by the municipality). Whenever choosing this possibility, the commercial tissue can decide either between a supply model of public law and a supply model of private law. The public law entities' option usually makes sense when small commercial points of sales are, jointly, trying to compete with more modern and larger opponents; on the contrary, modern and integrated commercial areas may choose a supply model of private law. Thus, private law operators, commonly designated as 3PL, third party logistic operators, will have the tendency of working logistic activities in more recent commercial tissues, resulting from the commercial offer concentration or whenever modern and big commercial complexes are present. Therefore, it is expected that the public law operators can develop their activity in a context where the commercial tissue still presents more traditional characteristics. It is evident that the role-played by public entities, entering the market to regulate and renovate city systems when commercial tissues are more conservative and atomised. The role of the public intervention/regulation assumes, then, the function of reference and of will conjugation
186 Logistics systems for sustainable cities around projects that, by themselves, do not gather enough critical mass to have competitive conditions when facing modern commercial models. Then if there is a larger propensity of public law operators entering, connected to the municipalities or to the central public administration, for determined business circumstances, the private law operators will tend to support the bigger commercial units and to operate in more evolved commercial tissues, being more open to outsourcing practices. Both operators end up recognising the need of professional management to obtain sustained results when working with logistics systems. This means that whether the commercial tissue is more modern and of greater dimension or more traditional and smaller, logistic competitiveness is always essential. So, even though there is not one best way of doing things, best practice suggests that there is a necessity to use logistic experts so that cost structures can be viable and accompanied by a core business focus of the various commercial players. Thus, it is not erroneous to preview the necessity of the public law logistic services operators in using partnerships with private law entities to decrease risks and to profit from skills they usually do not have. So the logistic activities rendered by public law operators may transfer risks to private entities when knowing that the partnerships with the commercial tissues do not have the same management skills or the most interesting competitiveness conditions. Furthermore, the financial availability of the public sectors is too poor when facing increased exigencies of the 'citizenship' demand. This way, it became possible for the public law entities to develop partnerships with entities of private law, even having in sight the regulation of the commercial tissues and the active participation of several players, bigger or smaller, more traditional or more modern. Regarding this perspective several options are possible: (1) Service Contracts (SC) - a simple supplying contract of the private law operator to the public law entity; (2) Design and Build Contracts (DBC) - contract where the private law operator plans and builds the supply infrastructure but where the public sector starts to manage that same infrastructure; (3) Design, Build & Operate Contracts (DBO) - contract where the private service operator plans, builds and operates the infrastructure and, eventually, also the transportation between the infrastructure and the selling points; although, the investment is provided by the public law entity; (4) Design, Build, Operate & Finance Contracts (DBOF) - contract where the private law services operator plans, builds, operates and finances the infrastructure and manages all the logistic system from the storage to the transportation (operates in substitution of the public sector, to whom revert all fixed assets at the end of the partnership's contract). So, the more complete the contract is and the partnership between the public law and private law entities the larger the risk that the public law entity transfers to the private law entity. Knowing that on one side, partnerships between logistic service providers and commercial entities can be exposed to a larger or smaller risk facing the existing supply conditions, the nature of the city system and its corresponding selling points and, on the other side, that the nature of the logistic services provider can also be more of public or of private law according
Systems theory, complexity and supply organizational models
187
to the nature of the specific city system and its particular commercial geography, more or less traditional, among others, one may conjugate these two perspectives in an attempt to foresee an answer to the original matrix where the city system is represented (Figure 1). In this situation one should analyse Figure 2.
Figure 2 Prospective Solution Matrix MODERN-DISPERSED cities (A type) can have better solutions private partnerships between the logistic services operators and commercial entities due to the evolution and to the modern pattern of the city's commercial geography. Without taking a large risk there is a tendency to establish private law partnerships (commercial spaces and logistic service providers) concerning the supply's optimisation. In this circumstance, public intervention does not have much regulative power once the small and more traditional selling points have low presence, in number, and the entropy is high, thus foreseeing a more modern commercial geography. MODERN- INCOMPLETE cities (B type) can also have better solutions in the area of the private logistic operators (in partnerships with commercial entities). Nevertheless, these partnerships may not recognise the expansion need and the coverage of other still incomplete commercial areas. Consequently, if support to the establishment of smaller units is not previewed in those places, namely in terms of supply, the urban space can become hostage of the private partnerships because areas not covered may present clear disadvantages in competitiveness. Incomplete urban spaces can have the tendency of creating large development drops, so that the less covered areas will originate harmony problems, commercial and even social integration difficulties. Clearly, in these cases, competitiveness is more complex and partnerships between small commercial entities and logistic service operators can incur in great risk. Public entities, in regulating, will have the tendency to immediately pass the risk to private logistic service operators. In any case, the risk remains and the tendency, even though indirectly, will tend to go always to the participation of private logistics service operators. TRADITIONAL-COMPLETE cities (C type) can have better solutions in the area of public law logistic operators, these are capable of joining various willingness and sensibilities coming to build an infrastructure project that can renovate traditional commercial tissues, releasing them in competitiveness terms. Being the partnerships' risk, by tendency, more reduced, due to a more traditional expression of the whole commercial tissue. There will be a tendency to
188 Logistics systems for sustainable cities recognise the lack of competitiveness condition's and the need of public intervention to regulate and return better conditions to more traditional commercial entities. Even though the partnership risk is low, there is nothing against risk transfer to the private law logistics operators. MIXED cities (D type) can have better solutions in the area of public logistics operators. Since niche areas are contributing to organise the commercial geography it is strongly beneficial to the entry of private law logistic operators, establishing partnerships with recent commercial spaces, thus creating new centrality's. In this case, the partnerships' risk of public logistic operators with the traditional commercial tissue will be low. There is enough critical mass to create good supply conditions when using public operators. Nevertheless, risk transfer is possible, namely when wanting to use private law logistic service operators.
CONCLUSIONS The paper has several important conclusions. Firstly, one has the validity of the systems theory application to the study of city systems. Secondly, the recognition that the presentation of a simple matrix of double entry can contemplate/resume some of the representations of urban spaces and its complexity, contributing to the drawing of possible solutions in terms of supply Thirdly, resulting from re-design of the city logistic system, the search for global solutions can become an interesting path to explore. Finally, the obvious conclusion that it would be interesting to pursue this conceptual work at the empirical level, analysing the city system and studying solutions that, in each circumstance, can be presented, so that one should validate, connect or rethink the current paths proposed.
REFERENCES Anderson Jr., E. G.; C. H. Fine and G. G. Parker (2000). Upstream volatility in the supply chain: The machine tool industry as a case study. Production and Operations Management, 9, 239-251. Barlas, Y. and A. Aksogan. (1996). Product diversification and quick response order strategies in supply chain management. Proceedings of The 14th Conference of the International System Dynamics Society, July 22-25, Cambridge, Massachusetts. Bertalanffy, L. V. (1968). General Systems Theory. Penguin Books, London. Cakravastia, A. and L. Diawati, (1999). Development of a System Dynamic Model to Diagnose the Logistic Chain Performance of Shipbuilding Industry in Indonesia. Proceedings of the 17th International Conference of the System Dynamics Society, July 20-23, Wellington, New Zealand. Carvalho, J. C. (2001). e-Business & e-Commerce -On& Offline. Edicoes Silabo, Lisbon. Christopher, M. (1992). Logistics and Supply Chain Management. Pitman Publishing, U.K., 1992. Checkland, P. and Scholes, J. (1999). Soft Systems Methodology in Action. John Wiley & Sons, 1999. Forrester, J. W. (1959). Industrial Dynamics: A Major Breakthrough for Decision Makers. Harvard Business Review, 36, 37-66. Forrester, J. W. (1961). Industrial Dynamics. MIT Press, Cambridge, Massachusetts. Forrester, J. W. (1989). The Beginning of System Dynamics. MIT System Dynamics Group Memo D-4165-1, Cambridge, Massachusetts.
Systems theory, complexity and supply organizational models
189
Fukunaga, Y.; Y. Takahashi; N. Tanaka; K. Nobuhide; T. Kojima and M. Morita (2000). System dynamics analysis of stability during non-equilibrium stage in physical distribution. Proceedings of the 18th International Conference of the System Dynamics Society, 6-10 August, Bergen, Norway. Gattorna, J.L. and D. W. Walters (1996). Managing the Supply Chain: A Strategic Perspective. MacMillan, London. Gharajedaghi, J. (1999). Systems Thinking - Managing Chaos and Complexity- Butterworth Heinmann. Groothedde, B. (2000). Dynamics In Spatial Logistic Chains. Proceedings of the 18th International Conference of the System Dynamics Society- 6-10 August, Bergen, Norway. Le Moigne, J.-L.. (1977). La Theorie du Systeme General: Theorie de la modelisation- Presses Universitaires de France, Paris. Lau, H. L.; V. Padmanabhan and S. Whang (1997). The Bullwhip Effect in Supply Chains. Sloan Management Review, Spring, 93-102. Shannon, C. E. and W. Weaver (1949). The Mathematical Theory of Communication- Urbana, University of Illinois Press. Simchi-Levi, D.; P. Kaminsky and E. Simchi-Levi (2000). Designing and Managing the Supply Chain - Concepts, Strategies, and Case Studies. Irwin - McGraw-Hill, Boston. Simon, H. A. (1981). The Sciences of the Artificial The MIT Press, 2 nd edition, Cambridge, Massachusetts. Simon, H. A. (1990). Prediction and prescription in systems modeling. Operations Research, 38, 7-14. Sobrero, M. and E. B. Roberts (2002). Strategic management of supplier-manufacturer relations in new product development. Research Policy, 31, 159-182. Sterman, J. D. (1991). A Sceptic's Guide to Computer Models. In: Managing a Nation: The Microcomputer Software Catalog (G. O. Barney ed.). Boulder, CO: Westview Press, pp. 209-229. Sterman, J. D. (2000). Business Dynamics: Systems Thinking and Modeling for a Complex World. Irwin - McGraw-Hill, Boston. Strohhecker, J. (2000). Supply Chain Management: Software Solutions Versus Policy Design. Proceedings of the 18th International Conference of the System Dynamics Society, 6-10 August, Bergen, Norway. Swaminathan, J.; S. F. Smith and N. M. Sadeh (1998). Modeling Supply Chain Dynamics: A Multiagent Approach. Decision Sciences, 29, 607-632.
This page is intentionally left blank
14
ASSESSING IMPACTS OF GREENHOUSE GAS ABATEMENT MEASURES ON URBAN FREIGHT
Leorey Marquez, CSIRO Australia, Australia Nariida Smith, CSIRO Australia, Australia David Kilsby, Kilsby Australia, Australia Mike Taylor, Transport Systems Centre, University of South Australia, Australia Rocco Zito, Transport Systems Centre, University of South Australia, Australia
ABSTRACT A study to investigate the sensitivity of urban freight patterns to various greenhouse abatement policy measures is underway with Metropolitan Sydney being used as the case study area due to the availability of detailed freight and passenger network level data and models at the New South Wales Transport Data Centre (TDC). The study is designed to build on methodologies under development by TDC to derive freight traffic due to total requirements for freight and relative requirements for categories of goods from actual or forecasted commodity flows and associated information. This paper describes the selection of candidate policy measures for investigation and presents the methodology and processes used in modelling their impacts on urban freight patterns. The discussion will focus on six scenarios which provide policy instruments for application to a 1996 base case. Some results of the modelling of these scenarios will then be presented and issues arising from the study discussed. Special attention will be given to the relative changes in travel characteristics and emissions brought about by these instruments.
INTRODUCTION The relative proportion of freight to passenger traffic in Australian cities is increasing. Although passenger demand is expected to plateau, there is no sign of that happening for freight. Thus the environmental impacts of freight traffic such as greenhouse gas (GHG) emissions are of increasing concern. In view of this, the Australian Greenhouse Office (AGO)
192 Logistics systems for sustainable cities has commissioned a study to investigate the sensitivity of urban freight patterns to various greenhouse gas abatement policy measures. The Sydney urban area is being used as the focus of the case study due to the availability of detailed freight and passenger network level data and models at the New South Wales Transport Data Centre (TDC). However the estimated impacts of changing freight vehicle operations across the urban area, via vehicle technologies, infrastructure improvements, logistic and land use changes are expected to be generally applicable to all Australian cities and potentially applicable to cities elsewhere. The study is designed to build on methodologies under development by TDC to derive freight traffic due to total requirements for freight and relative requirements for categories of goods from actual or forecasted commodity flows and associated information. These are to be used in conjunction with a model of passenger travel, which estimates the passenger traffic on the urban road network. Passenger traffic information is needed since the performance of freight traffic is dependent on the volume of passenger traffic on the network. However while the TDC passenger travel model is behaviourally based, the freight estimation, in common with numbers of other freight models, is being based on commodity movements. In essence, the model links the demand for different commodities to be moved from A to B with the usual ways and means of getting the freight there. Thus the TDC model will predict changes in traffic on the network due to changes in the need for different types of freight but not changes in the usual means of getting there. This study defines a process for investigating the impacts of policy intervention on ways and means of moving freight. This is particularly important as policies for GHG abatement seek better emissions outcomes without hindering the supply of goods.
Figure 1 Process for applying policy measures
Assessing impacts ofGHG abatement measures on urban freight
193
Figure 1 shows a diagram of the modelling process by which the Transport Data Centre (TDC) synthesises a trip matrix of commercial vehicle movements for Sydney. The process is designed to convert a base year snapshot of commodity flows into an estimate of commercial vehicle movement. The base year is currently 1996, and the flows are based on the Freightlnfo database of FDF Pty Ltd, a Melbourne company specialising in freight flow estimation. This database will be updated, but will remain retrospective, when later data is released by FDF. There are some exclusions from the FDF database, of which the most notable is waste disposal. Therefore the eventual trip matrix will not include any estimate of garbage truck movements. The volumes of different vehicle and fuel types on a link-by-link basis are determined as part of the outputs of Sydney's Strategic Traffic Model (STM) / CTS (Commercial Transport Study) for each policy alternative to be tested. These outputs essentially define the transport task by vehicle type. The four types of vehicles used in the model are passenger vehicles (PV), light commercial vehicles (LCV), and rigid trucks (RT) together with articulated trucks (AT). One of the limitations of this method of classification of outputs is that none of the vehicle categories are further disaggregated by fuel type. The fuel used by a vehicle obviously makes a significant contribution to the characteristics and quantities of emissions.
Figure 2 Thematic map of LCV link volumes in 1996 base case Intervention in the TDC modelling process will predict the effect on 1996 commercial vehicle movements if the various policy measures to be tested had been implemented in 1996. This is referred to as the Base Case. Thus, for the road system, the base network used is a subset of the Sydney road network that existed in 1996 consisting of around 15,000 one-way links and excluding local roads. Network additions, which have been opened since 1996, including the M2 and the Eastern Distributor tunnel, were also not included. Figure 2 shows a thematic map of this limited network.
194
Logistics systems for sustainable cities
The strong interactions between road freight and other road users and the impacts of policies on these are modelled explicitly by combining commercial trip tables with car trip tables, and assigning both to the same road network. Multi-class assignment techniques allow reporting on each vehicle type separately despite the interaction between them. The traffic on the network is derived from trip tables for travel between origin and destination traffic zones. Link volumes are calculated for three categories of commercial vehicles: light commercial vehicles (LCVs), rigid trucks (RTs) and articulated trucks (ATs). VKT (vehicle kilometres travelled) and speeds are estimated for both the 1996 base case year and change scenarios. Base passenger car traffic emissions are then added to the freight scenario emissions resulting in sets of emissions within time of day periods.
POLICY INSTRUMENTS AND ASSESSMENT The first stage of this study, conducted in April to July 2002, used a literature search and expert advice to develop a qualitative understanding of the urban freight task, and investigate a range of potential policy measures. A modelling framework was designed to test the measures. The study concluded with a workshop for the project steering committee and other stakeholders. This led to an agreed set of measures to be tested (Table 1) by the methods proposed.
Category Infrastructure Measures Vehicle Movement Measures Planning and Land Use Measures Vehicle Measures
Table 1 Proposed Policy Instruments Instrument • General reduction in congestion • Improved traffic management • An orbital road project • • •
Higher (and lower) load factors Real time traffic information Industry relocation as a result of an orbital road
•
'Best Practice' truck fleet fuel efficiency
Note, that the selected policy instruments actually refer to a desired state perhaps achieved from a set of unspecified initiatives. Thus, 'general reduction in congestion' may be brought about by a combination of various traffic initiatives, taxation policies, and infrastructure projects. The focus of the study is not on the individual initiatives that can help bring about a 'general reduction in congestion', but rather on the impact that this instrument, once it has been achieved, will have on urban freight. This approach keeps the study from being bogged down with the complexities of assessing each of a myriad of individual initiatives available. Various performance measures are calculated in order to obtain a qualitative assessment of the magnitude of the potential effects of the policy instruments. The assessments are based on the propensity to: • Reduce the number of vehicles required to service the freight task; • Reduce the required number of trips; • Reduce the trip length; • Reduce VKT for the given freight task; • Reduce fuel consumption, and • Reduce the rate of non-GHG emissions per litre of fuel consumed.
Assessing impacts ofGHG abatement measures on urban freight
195
For each measure, the possible performance is rated against the following criteria: • Relative size of the GHG effect; • Market size affected; • Overall GHG emissions; and • Overall Non-GHG emissions.
BASE CASE SCENARIO All the chosen policies represent deviations from the 1996 business-as-usual settings. Therefore the first requirement was to model the 1996 base case to obtain the benchmark against which all policy variations are compared. As noted above, the purpose of modelling is to obtain strategic level estimates of policy impacts and compare options rather than produce absolute values for forecasting purposes. Thus modelled estimates of commercial vehicle and car trip-making patterns for 1996 should be regarded as the best available with current data and models. Their accuracy was not an issue for this study and therefore no validation was undertaken. The question was rather, given that these matrices described complex movement patterns in detail, what effect on those patterns (and hence emissions) would the selected policies have? Table 2 shows the four weekday time periods used in the route assignment, the length of each period and the factor used to convert to a representative 2-hour period. The conversion was necessary because the descriptions of link capacities in the network relate to 2-hour capacity. Hence it was necessary to convert trips in each of the four time periods to their two-hour equivalents for assignment, and then to reverse the conversion to obtain the results for the full time period. Table 2 Durations of weekday time periods and conversions to 2-hour period Time period Conversion Duration Factor 1 AM Peak (7am-9am) 2 hours Business Hours (9am-3pm) 6 hours 0.33 PM Peak (3pm - 6pm) 3 hours 0.67 Evenings (6pm - 7am) 13 hours 0.25 It was also necessary to convert the larger commercial vehicles to "passenger car unit" equivalents (pcu's) before assignment, because the network capacity and behaviour (via the volume-delay curves used in assignments) are based on pcu's rather than vehicles generically. As with the time periods, the reverse conversion is required after assignment to turn the results back into commercial vehicle units. Following common modelling practice, the pcu equivalents used for the three types of commercial vehicles are: • Light commercial vehicles = 1 pcu • Rigid trucks = 2 pcu's • Articulated trucks = 3 pcu's Table 3 shows the number of weekday trips made by vehicle type and (expanded) time period, for the 1996 base case. The table also shows the proportion of vehicles in each class travelling at each time of day. For cars, these were obtained by a complex system of factoring based on
196 Logistics systems for sustainable cities data collected for the ongoing Household Travel Survey. Thus, passenger vehicles or cars account for more than 92% of all daily trips with 36% of all trips made during business hours. For commercial vehicles, the corresponding splits were obtained from the Commercial Vehicle Survey of 1992. Table 3 Percentage of weekday trips (000) in Sydney by vehicle type and time of day Vehicle type AM Peak Business PM Peak Evenings 24 hours Hours Cars 1,146 (16%) 2,455 (35%) 1,734 7,113 1,778 (100%) (24%) (25%) Light CVs 74(17%) 215 (48%) 87 (20%) 65 (15%) 443 (100%) Rigid trucks 26 (19%) 73 (53%) 19 (14%) 20 (14%) 138 (100%) Articulated trucks 5 (22%) 5(19%) 13 (49%) 3 (10%) 25 (100%) All vehicles 1,825 1,251 (16%) 2,755 (36%) 1,888 7,720 (100%) (24%) (24%)
POLICY SCENARIOS The seven policy scenarios and the base case are evaluated in line with the procedure shown in Figure 1. As indicated, the TDC process is designed to convert a base year snapshot of commodity flows into an estimate of commercial vehicle movement. Our estimation is currently based on 1996 tables and addresses combined freight and passenger traffic in four time of day periods - morning peak, midday, afternoon peak and evening/night. With this methodology, it is possible to model the effects of "infrastructure measures" by adding the commercial vehicle trip pattern estimates to those of the car driver trip patterns produced by the passenger travel model. Changes in congestion are modelled via an appropriate reduction of passenger traffic. The loading factors are addressed by a reduction of trips due to increased loads and via use of larger vehicles producing changes in relative numbers of trips by the vehicle classes of light commercial vehicles, rigid truck and articulated trucks. In both cases the proportion of changed trips is estimated from the opportunities for higher loads in commodity classes and for use of larger vehicles. Unfortunately, the converse situation of more frequent trips at lower loads due to increases in "just in time" deliveries was not tested due to time constraints. Changes brought about by planning measures are effected through a redistribution of the population of workers of industries deemed most likely to be susceptible to the measures. This results in a revised trip distribution matrix using new locations for the selected industries. Fuel efficiency changes are tested in the emission estimation procedure which follows the network traffic information. The detailed network speed estimates by time of day allow congestion sensitive estimates of the variation in impacts of better fuel efficiency across the urban area. The implementation of each of the policy scenarios in the transport model is briefly summarised as follows. Scenario 1: Improved Fuel Efficiency
Assessing impacts of GHG abatement measures on urban freight
197
For the purposes of modelling, we assumed that improved fuel efficiencies in the commercial vehicle fleet would only effect the emission outcomes, not the amount of travel. The emissions are calculated from 1996 base case travel patterns but using the lower fuel consumption coefficients. The elasticity figures from stage 1 justify the assumption that better fuel efficiencies and hence lower fuel costs are not likely to induce significant amounts of extra freight travel. Thus, this scenario did not utilise the transport model. Scenario 2: Lower Congestion This scenario assumes that there are a number of ways to reduce congestion ranging from the encouragement of public transport to enforcing parking restrictions to improve free flow. The outcome of such policies is then modelled by assuming a 15% reduction in congestion, due to passenger vehicles, in the AM and PM peaks. Information from stage 1 of the study suggests that a 15% reduction in congestion would be reasonable. Scenario 3: Better Traffic Management Again this scenario encompasses a number of strategies to better manage traffic flow on the network, from removing bottlenecks to sophisticated intelligent transport systems for traffic management. The performance of roads in the network under traffic is governed by a relationship known as the Volume-Delay Function derived from the capacity of the road and the numbers of vehicles using it. This scenario adjusts this function so that at saturation point each arterial road would be operating at 3kph faster than previously, and 10% would be added to the traffic capacity of the road. Arterial roads potentially benefit most from improved traffic flow management. Freeways/motorways are more likely to have systems already in place and sub-arterial/local roads usually do not reach their capacity. Scenario 4: Higher Load Factors In this scenario, load factors are increased for commercial vehicles. This can be achieved by higher loads per vehicle, or the use of a larger vehicle either in the same class or a larger class. All result in fewer vehicle trips to move a given quantity of goods around the city. The scope for this sort of change varies from industry to industry. The 1996 commercial vehicle trip matrices were factored to take into account feasible changes in such practices, given the location of industries and the types of vehicles involved. The industry mix in each area is weighted by an estimated likely percentage trip change in each industry to estimate potential changes in each travel zone.
Scenario 5: Real-time Traffic Information Provision of real-time traffic information to improve network performance was modelled using the same Volume Delay Function as used for traffic management with the information
198
Logistics systems for sustainable cities
provision restricted to the principal arterial routes to the Sydney and Parramatta CBD (Central Business District) and the main ring routes.
Scenario 6: Infrastructure Improvement This policy test assessed what difference the Sydney Orbital Route, expected to be completed in 2006, would make if it had been present in 1996 and if the origins and destinations for commercial and private vehicles trips remained as they were without the Orbital. Figure 3 provides a sketch map of the western route in relation to other links in the Sydney orbital network.
Figure 3 The proposed Western Sydney Orbital Route
Scenario 7: Infrastructure Improvements with Land Use Change The Sydney Orbital Route improves access from Western Sydney. This scenario evaluates the impact if, in addition to the new orbital infrastructure in the previous scenario, businesses in Inner Sydney moved west closer to the orbital route. 10% of manufacturing jobs were moved to travel zones near the route and the impact of the estimated loss and gain of 2.2 freight trips per day due to each of those jobs was estimated. Table 4 summarises the implementation approach to the selected policies adopted by the transport modelling component of the study.
TRANSPORT MODELLING RESULTS Seven runs were performed to produce the link volumes for a 24-hour period representing the travel patterns under the given scenarios. The runs and their corresponding scenarios were performed as follows: • Run 1: 1996 base case (Also used for Improved Fuel Consumption)
Assessing impacts ofGHG abatement measures on urban freight • • • • • •
Run 2: Run 3: Run 4: Run 5: Run 6: Run 7:
199
General reduction in congestion Improved traffic management Higher load factors Real time traffic information Infrastructure improvement - the Sydney Orbital The Sydney Orbital with industrial relocation
Table 4 Transport modelling approach to selected policies Policy (Scenario) Transport modelling approach Improved fuel Travel patterns from 1996 base case. Calculate emissions consumption (Seen 1) directly from lower fuel consumption figures. Lower congestion (Seen 2)
15% reduction in car use in AM and PM peaks
Better traffic management (Seen 3)
Volume-delay functions on arterial roads modified to increase speed at saturation by 3 kph and to add 5% to capacity.
Higher load factors (Seen 4)
Commercial vehicle trip matrices modified to give same quantity of goods being moved between same places but with higher load factors and some transfer of goods to larger vehicles.
Real-time traffic information (Seen 5)
Volume-delay functions on major approaches to CBD and Parramatta and on major orbital routes modified as per "better traffic management".
Infrastructure improvement (Seen 6)
Sydney Orbital route at freeway standard added to 1996 road network for Sydney.
Infrastructure improvement with land use and distributional feedbacks (Seen 7)
Sydney orbital route at freeway standard added to 1996 road network for Sydney; westward shift of employment assumed; modified CV patterns as a result
Again, base case (Run 1) results were used as a reference for assessing the relative impact of the six policy measures. Figure 2, shown earlier, provides a thematic map of the volumes of LCVs on the road network for the midday period of the base case. The graduations in this map have been split up into eight classes with different colours and line widths to illustrate the spatial distribution of the volumes. These thematic maps can be produced, if required, for any emission, vehicle and fuel type combination used in the emissions model. As validation, the base case results have been compared to results using simplified emissions factors as given by the Australian Greenhouse Office (AGO, 1998) and Environment Australia (2001). All the results showed strong correlation, and hence provide a firm basis for the analysis of the other policy options. The results of applying each of the policy scenarios in the transport model are briefly presented as follows. Lower congestion
200
Logistics systems for sustainable cities
Lowering peak congestion reduces both peak VKT and VHT. The percentage reduction in VHT exceeds that of VKT and average travel speeds increase. This only occurs in peak periods and the 24-hour performance is therefore watered down. The majority of commercial vehicle movement takes place outside the peaks. Better traffic management Improvement of the performance of arterial roads improves traffic flow thus reducing VHT and increasing average speed. But the overall effect is small, because the better-performing roads tend to attract more traffic, which slows them down again. Higher load factors A move to higher load factors and load consolidation produces a large net decline in VKT by commercial vehicles and hence is likely to reduce emissions. The use of larger trucks would reduce the number of trips but there is very little change in operating speed, since VHT declines in roughly the same proportion as VKT. Real-time traffic information This increases the performance of the principal arterial and orbital routes. It has practically no effect on commercial vehicle VKT but VHT decreases slightly and hence higher operating speeds are achieved (for light CV's and rigid trucks). Effects of infrastructure improvement If trip patterns did not change, the addition of the Sydney Orbital to Sydney's road infrastructure in 1996 would have encouraged longer but faster trips by both cars and commercial vehicles, with the result that VKT would go up, VHT would go down and average travel speed would increase. Effects of infrastructure improvement with land use change Relocation of some freight-generating employment from inner areas to Western Sydney actually increased commercial vehicle activity because some of the displaced movement would still have the docks and central industrial areas as its destination pattern. Some increase may be a result of modelling assumptions but it is also likely that larger scale land use changes such as new terminals are needed for freight travel.
EMISSIONS MODEL On completion of the transport model runs, data was passed on to the Transport Systems Centre (TSC) which developed the model for calculating total emissions for each link in the
Assessing impacts ofGHG abatement measures on urban freight
201
network. The emission model covers 12 different types of greenhouse gas and air quality emissions as well as the four vehicle types (passenger vehicles, light commercial vehicles, rigid trucks and articulated truck) used in the study. The emissions species considered include CO, CO 2 , VOC, NOx, N 2 O , SO2, CH4, PM, 0 , Benzene and l,3Butadiene. Each vehicle type has been further disaggregated into three different fuel types: petrol, diesel and liquefied petroleum gas (LPG). The outputs from the transport models however do not provide information on the numbers of the different vehicles by fuel type. These vehicle-fuel numbers were obtained by multiplying the appropriate vehicle number by the proportion of the fleet that makes up the particular fuel type. The proportions of vehicles by corresponding fuel types were obtained from the 1997-1998 ABS Small Area Motor Vehicle Data (ABS, 1999). In all, 132 combinations were investigated for use in this study. Emission estimates were not derived for LPG-fuelled articulated vehicles since ABS (1999) showed that there were only three of this type of vehicle-fuel combination in the Sydney metropolitan area and hence would have no significant effect on total emissions produced. The emission model used parameters from four information sources. The first source of information was the European emissions inventory guidebook (European Environment Agency, 2002). The guidebook contains emission and fuel consumption functions for the four vehicle types listed. Most of these functions are sensitive to varying link average speeds and differing vehicle loads. This is an important requirement since many of the policy scenarios that are being tested require this degree of sensitivity. However, since the study is concerned with the Australian vehicle fleet, a method of scaling these functions was developed to allow these European models to reflect the emissions characteristics of the Australian vehicle fleet. Scaling factors were derived from the second and third information sources namely the Australian Greenhouse Office (AGO, 1998) and Environment Australia (2000). The fourth information source was the Apelbaum Consulting Group Transport Facts publication (Apelbaum Consulting Group, 2001), which was used to derive fuel consumption values for the 1996 vehicle fleet. The information sources used were considered to be the best sources of data for use in this study.
DISCUSSION OF PARTIAL RESULTS The result of the runs will be presented in detail in the projects final report. This section presents representative results for two of the emissions: carbon dioxide (CO2) and particulates (PM 10 ).
CO2 emissions Figure 4 displays a comparison of the daily CO 2 emissions from each type of commercial vehicle for the seven runs. The chart shows that the emissions from Run 4 (Higher Load Factors) present significant reductions from those of the base case (Runl) especially for LCV's and rigid trucks. Run 2 (General reduction in congestion) also showed improved results. On the other hand, the results of Run 7 show increased emissions from the base case, with emission levels doubling for articulated trucks. However, these results should be viewed in the context of overall vehicle emissions. Total daily CO2 emissions from passenger vehicles are about ten times those of freight vehicles, as shown in Figure 5. This means that policy measures that impact passenger vehicles will
202 Logistics systems for sustainable cities produce greater reductions than those targeting freight vehicles. Thus, Run 2 results provide more significant reductions than Run 4. In fact, even Run 7 results, when viewed from an overall perspective, offer long-term attraction in terms of reducing car emissions in spite of potential increases in freight emissions.
Figure 4 24-hour CO 2 emissions from freight vehicles only
Figure 5 24-hour CO 2 emissions from all vehicles Table 5 further illustrates the implications of this imbalance. As pointed earlier, Scenario 4 can significantly reduce CO 2 emissions from freight vehicles, up to 17.3%. On the other hand, the application of Scenario 7 may increase freight emissions by 21.4%. However, since freight vehicles only account for 16.3% of all vehicle CO 2 emissions, and all scenarios reduce emissions from passenger vehicles, all scenarios offer emissions reduction benefits overall. Scenarios 3, 4, 5 and 7 appear to offer the same 3% level of reductions over all traffic while
Assessing impacts of GHG abatement measures on urban freight
203
Scenario 6 offers a reduction rate of 6.6%. Scenario 2 appears to have the most significant reduction potential at 17%. Table 5 CO2 emission deviations from base case by scenario Freight Traffic Only ALL Traffic Policy Instrument (Scenario) -4.8 % -17.1 % Lower congestion (Seen 2) -3.3 % -2.8 % Better traffic management (Seen 3) -17.3 % -3.4 % Higher load factors (Seen 4) -3.2 % -3.2 % Real-time traffic information (Seen 5) -1.0% -6.6 % Infrastructure improvement (Seen 6) +21.4% -2.6 % Infrastructure improvement with land use and distributional feedbacks (Seen 7) Base Case: Freight vehicles contribute 16.3% of total CO2
PM10 emissions In contrast with carbon dioxide emissions, particulate emissions have more variable sources. Figure 6 shows that rigid trucks contribute almost as much emissions as passenger cars with articulated trucks and LCV's following close behind. This results in freight vehicles accounting for 51.5% of all PM10 vehicle emissions. According to Table 6, Scenario 4 again provides the best potential for reducing PMJO emissions from freight at about 17.7%. Scenario 2 comes second at 6.6%. Scenario 7 again creates the opposite effect with increases in freight emissions of up to 20.1%. Since the proportion of passenger vehicles is considerably smaller, the overall effect of Scenario 7 continues to point to increases of 6.8%. Scenarios 2 and 4 again provide the best options for overall reduction, at 12.9% and 9.5% respectively.
Figure 6 24-hour PMio emissions from all vehicles
204
Logistics systems for sustainable cities
Table 6 PMio emission deviations from base case by scenario Policy Instrument (Scenario) ALL Traffic Freight Traffic Only -12.9 % -6.6 % Lower congestion (Seen 2) -3.8 % -4.1 % Better traffic management (Seen 3) -9.5 % -17.7 % Higher load factors (Seen 4) -3.7 % -4.0 % Real-time traffic information (Seen 5) -4.6 % -1.6 % Infrastructure improvement (Seen 6) +6.8 % +20.1 % Infrastructure improvement with land use and distributional feedbacks (Seen 7) Base Case: Freight vehicles contribute 51.5% of total PMio
CONCLUSION Table 7 below summarises the effect of the policy instruments on the performance measures chosen for assessment. All instruments have the potential to reduce VHT, trip length as well as emissions. The use of 'higher load factors' has positive implications across the board, including reductions in the number of vehicles, reductions in the number of trips, and significant reductions in freight emissions. However, this scenario is contrary to industry trends towards lower load factors and "just in time" deliveries, which are likely to increase emissions. In contrast, the positive effects of 'infrastructure improvement with land use changes and feedbacks' were limited. The principal effect assumed as a result of allowing land use to respond to the introduction of the Sydney Orbital was a relocation of some freight-generating employment from inner areas to western Sydney. When this was input to the model, the prediction was that this would actually increase commercial vehicle activity if the same destinations were to be served. This is because some of the displaced movement would still have the docks and central industrial areas as its destination pattern. The impact was most noticeable for articulated trucks, whose VKT was predicted to increase by 27% per day. Table 8 presents an assessment of the size of the impacts of the policy instruments based on four criteria. It is clear that all instruments present significant effects relative to the freight task with 'General reduced congestion' and 'Higher load factors' having strong positive effects while 'Orbital with land use change' pose strong negative effects. Overall, the transport and emission modelling results suggest that 'General reduced congestion' presents the widest coverage and strongest effects in terms of market size, GHG emissions and non-GHG emissions.
Assessing impacts of GHG abatement measures on urban freight Table 7 Performance measures of policy instrument Type of Effect Policy Instrument Reduce Reduce Reduce Reduce VKTs No. of No. of Trip Length Vehicles Trips General reduced YES YES congestion Better traffic YES management YES YES Higher load factors YES YES Real-time traffic „ YES information YES YES Sydney orbital road YES Orbital with land use YES change
205
effects Reduce Reduce Fuel Emissions Use per Litre
Table 8 Impact assessment of policy instruments Size of Effect Market Overall Relative GHG Size Policy Instrument Size of GHG Effect Affected Emissions General reduced ,/,/,/v • • • •••• congestion Improved traffic . •••• management V Higher load factors •/•/>/•/ Real-time traffic ,, V information • Sydney orbital road •/•/ Orbital with land use . . V V V V change
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES YES
Overall Non-GHG Emissions ••• V V
REFERENCES ABS (1999). TranStats Small Area Motor Vehicle Data 1997-98, Report No. 9312.0.30.001, Australian Bureau of Statistics, Commonwealth of Australia. AGO (1998). Energy Workbookfor Transport (Mobile Sources) 3.1, Australian Greenhouse Office, Commonwealth of Australia. Apelbaum Consulting Group (2001). Australian Transport Facts 1998, Report No. 0642721726, The Centre for Transport Energy and the Environment. Environment Australia (2000). National Pollutant Inventory Emission Estimation Manual for Combustion Engines Version 2.1, Commonwealth of Australia. Environment Australia (2001). National Pollutant Inventory Summary Report of Second Year Data 1999 - 2000, Report No. 0 642 54694 0, Commonwealth of Australia. European Environment Agency (2002). Emissions Inventory Guidebook, Report No. 3, European Environment Agency.
This page is intentionally left blank
15
THE ENVIRONMENTAL ASSESSMENT OF URBAN GOODS MOVEMENT Erwan Segalou, PhD student, Laboratoire d'Economie des Transports, France Christian Ambrosini, Senior lecturer, Laboratoire d'Economie des Transports, France Jean-Louis Routhier, Research engineer, Laboratoire d'Economie des Transports, France
ABSTRACT This paper presents the implementation of a thorough physical environmental assessment of urban goods movement (UGM). It is aimed at local decision makers concerned with sustainable development issues in urban planning. After a description of the new stakes for urban planners, we specify the data processing method and the sequence of the models used as well as the main results in three French cities: Bordeaux (750,000 inhabitants), Dijon (240,000 inhabitants) and Marseilles (1,050,000 inhabitants).
INTRODUCTION In France, the development of urban environmental assessment is recent. Most of the research studies about it are concerned with either the whole transport sector on a national scale, with no distinct spatial levels, or consider only the impacts of private cars traffic (Gallez and Hivert, 1998). The necessary data bases and models able to describe the urban goods vehicles traffic in order to carry out a complete assessment were almost non existent in many countries (Ambrosini and Routhier, 2003). In many cases only heavy goods vehicle (HGV) flows are identified. However, HGV flows represent only one part of urban goods movement (UGM). In France, apart from shopping trips, HGV (> 3.51) account for about 50% of the UGM vehiclekilometres (Routhier et al, 1996-99), whereas a third of light goods vehicles (LGVs) that weight below 3.5 tonnes also carry goods on a regular basis (SES, 1999). It is thus essential to accurately identify the contribution of UGM in traffic generation, with the help of a relevant knowledge database. The method presented in this paper in an extension of the data acquisition process carried out in France for several years towards a complete environmental assessment of UGM. It is a policy oriented model which aims at helping local decision making to assess the
208
Logistics systems for sustainable cities
contribution of the various actors within city logistics in relation to energy consumption, congestion and environmental concerns (pollution, greenhouse gas and noise).
STAKES In Europe, in the 1970s, local authorities focussed on the nuisance generated by heavy vehicles. It was primarily a question of limiting the local environmental nuisance: noise, visual impact and road safety. Various measures such as building by-passes and ring roads led to an exclusion of the through-traffic of heavy vehicles from the city. Increasingly restrictive regulations have limited lorry traffic in urban centres concerning the capacity and the size of vehicles, the allotted slots for deliveries, and the nature of goods (in particular hazardous substances). These short term local measures against UGM, have neglected the need to provide these 'protected' areas with goods and services. Consequently, perverse effects appeared: local authorities encouraged logistic platforms and shopping centres to relocate around large interchanges, at distances increasingly far away from urban centres and consumption areas. These relocations induced not only an increase in average trip length for delivery vehicles but also an increase in vehicle shopping trips. Thus today, several phenomena interact and have led to a fast degradation of the conditions of urban road traffic regarding the environment: (a) more significant distances between points of distribution and consumption; (b) road traffic increasingly constrained within a limited space, saturated by private car traffic; (c) an increase in commercial traffic induced by strong changes which are dependent on trends such as globalisation and intense competition (spatial division of labour, storage surface decrease and just in time deliveries). In France, peak hour congestion previously existed only in dense urban centres. But more and more large peripheral highways tend to be overrun with congestion. Thus areas and populations affected by direct pollution are increasing. At present, the development of road infrastructure in densely populated areas seems to have reached its limits. At the same time, transport demand keeps on increasing, citizens are more and more concerned with the quality of their environment and countries that are signatories of the Kyoto treaty are involved in decreasing greenhouse gases. In such a context, the overall stakes become more significant. This is why a complete environmental assessment considering the whole urban supply process has become more relevant than simply measuring the impact of heavy vehicles.
METHODOLOGY Impacts of urban transport on the environment can be observed at two levels: on a local scale: air pollution, noise, level of safety, water pollution, ground pollution and sharing of public space (roadway system and buildings); on a global scale, two major impacts: fossil energy consumption and greenhouse gas production.
The environmental assessment of urban goods movement
209
Our definition of the UGM takes into account the entirety of goods movement carried out in a town. This is why in addition to the deliveries and pick-ups to the economic or business premises we add purchasing trips, which represent the final link of the household supply chain, and freight vehicles movements related to urban management such as building materials, transport of waste, removals, etc. The traffic generated by these three types of activities constitute the total UGM. It is compared with two other types of flow: the freight through traffic (FTT) and the other motorised private or professional car trips. This typology has the advantage of being built on the basis of comprehensive data that is also generally available. Two main models are applied successively, as shown in Figure 1: 1) a two step traffic model: (i) generation-distribution of the traffic (five segments); (ii) traffic assignment on the network with six types of vehicles, according to energy consumption: private vehicles (PV), light goods vehicles (< 3.5 t), goods vehicles (3.5 to 7.5 t), GV (7.5 to 16 t), GV (16 to 321) and heavy goods vehicles (> 321). 2) a two step environmental model: (i) a model of pollutants emissions, greenhouse gases and noise; (ii) a model of pollutants concentration in the atmosphere.
Figure 1 The complete UGM environmental model
210 Logistics systems for sustainable cities Traffic model The Vehicles Movement Generated by the Deliveries and Pick-ups The whole of the goods flowing between premises (industry, trade, tertiary sectors, public or private services) was estimated from specific surveys carried out in France in three pilot cities (Routhier et al 1996-99). The method combines a premises survey (i.e. a survey of business addresses) and a survey of drivers. Premises survey: the observation unit is the operation of loading and unloading described by a sample of 4,500 premises spread through the three cities. The sample is stratified (a priori, then a posteriori) according to 45 main types of activities which have a common logistic organisation. These surveys allow: (a) to describe in a log book over a one week period the whole of the movement of goods vehicles generated by the premise; (b) to know the environment of the premise (roadway system, storage area and parking facilities); (c) to describe the delivery conditions (duration, nature and packaging of goods, type of vehicle and handling equipment used); (d) to specify the nature of the operation (for hire or own account, multi-drop or single delivery). Thus these surveys make it possible to describe the main determinants of the generation of the pick-ups/deliveries. Table 1 Numbers of vehicle movements per week in the three cities Premises Number of: Inhabitants Movements per week Bordeaux 750,000 40,466 310,000 Dijon 240,000 11,569 94,000 Marseilles 1,050,000 57,967 380,000 A total of 2,200 drivers were interviewed (drivers survey) in order to describe precisely the type and the routes of the vehicles, the modes of organisation of the runs (single delivery and multi-drop rounds of various sizes), the speed (time and place) of the vehicle, as well as the various activities served. These surveys made it possible to define the network occupancy by the various types of vehicles in urban areas. Then vehicle flows are directly assigned: the traffic is assigned along the routes after a correction according to the number of pick-ups/deliveries generated by the premises served in each area of the town. Table 2 Daily distance covered by the goods delivery vehicles inside the three cities Distance (in km) Rigid trucks Trailer-trucks <3.5t Bordeaux 213,784 160,240 110,379 Dijon 41,739 32,146 13,904 Marseilles 340,025 229,721 159,681 Private shopping vehicle trips Private shopping trips constitute the second UGM traffic segment taken into account. These trips form the last link of the distribution chain, from producers to final consumers.
The environmental assessment of urban goods movement
211
In France, the only data easily and directly available to study private shopping trips are derived from household urban travel surveys (Segalou, 1999). These have the advantage of being implemented with a standardised methodology in most French urban areas (Certu, 1998). French household urban travel surveys are face-to-face interviews in order to collect: (i) socio-economic data on household and on every person in the household; (ii) information on all trips carried out the day before the survey day by each member of the household: trip purpose, transport mode, origin/destination, departure time and duration, etc. On the basis of these data, it is possible to quantify and locate private shopping trips and to measure modal split. However, from these surveys it is not possible to estimate directly the private shopping trip matrix, required by the environmental model. Indeed, the information collected is often too vague to allow analysis on a detailed zoning system. In view of the sampling fraction, many origin/destination links have only limited data or are not statistically representative. These difficulties lead to a need for modelling in order to build the private shopping vehicle trip matrix. Then the results of household urban travel surveys were analysed and aggregated to define behaviour patterns relating to zone features of the study area. On the basis of the highlighted relations, calibrated using household urban travel surveys data, it is possible to determine the origin/destination matrix (Routhier et al, 2002). Consequently, these trips can be assigned onto the appropriate network and be taken into account by the environmental model. The activity which motivates the trips being considered is defined as a purchasing activity. This definition is more restrictive than simply going shopping where in the case of some trips no purchasing takes place. In practice, 'private shopping trips' corresponds to the following modelled trips: (i) trips whose origin is a store after a purchase. Indeed, it is after a purchase that the consumers carry goods in their vehicle. These trips are in particular modelled by taking into account the number of employees in two different types of stores (400 m2 < sales area < 2500 m2; sales area > 2500 m2) and the number of stores whose sales area is lower than 400 m2; (ii) trips from home to stores when the purchase is the only purpose of a home-based trip chain. In this case, an outward and return journey can be assigned to a purchasing activity. Urban management traffic In most surveys, some flows are not clearly identified. Regarded as negligible and fluctuating in time or space, they are not subject to thorough surveys. They come from overall estimations calculated on the basis of rough ratios (Table 3). The annual average mileage depends on the size of the city and on the location of waste sites, quarries and cement works. The location of these flows is not well known. That is why they are assigned in the same way as the flows generated by the deliveries and pick-ups.
212 Logistics systems for sustainable cities Table 3 Distance covered in the Bordeaux urban area by the urban management traffic Activity
Method of computation
Public works Demolition and building sites
500 trucks per week for 100,000 inhabitants 130 heavy vehicles per 1,000 m2, numerous light commercial vehicles Annual mileage of the specific vehicles
Networks management* and public services Household refuse and industrial Annual mileage of the garbage trucks waste collection Removals 10% of households, stores and firms move each year. Annual mileage of the specific vehicles in Post Office providing the urban area • (cleansing, water, gas, electricity, cable and telephony).
Km per inhabitant each year (% HV) 12.1
(81%)
17.9
(36%)
23.1
(48%)
3.9
(22%)
8.6
(40%)
Long distance and through traffic This includes flows of the large regional or international platforms on the outskirts of the metropolitan areas, as well as the interregional freight vehicles which cross or skirt the urban area. These flows have a different impact according to the shape and the geography of the town. "Cordon line" surveys provide the required data. They make it possible to count all the vehicles of more than 3.5 tonnes as well as the light commercial goods vehicles. Private individuals trips (other than shopping) They consist of private car trips taking place inside the town, between the town and its surrounding area and through the urban area. These flows are obtained by considering the difference between the total traffic modelled by the usual four step traffic models (assignment according to the Wardrop principle, applied to the whole of the flows) and the traffic flows of heavy goods vehicles, light commercial goods vehicles and shopping trips calculated as described in the previous sections. In the three cities, the traffic (vehicle-km) generated by the deliveries and the urban management in one day is about 5 to 6% of all traffic, i.e. 11% to 12% (car-unit-km). Shopping trips varies from 7% to 13% according to the city. UGM therefore varies from 13% to 19% (vehicle-km) but accounts for about 18% to 25% (car-unit-km). Table 4 Average daily traffic flows in the three cities (Aria Technologies, Systems Consult - Polydrom / 2000-2001) Traffic measured in daily vehicle-km Segment of daily traffic on the study area Bordeaux Dijon Marseilles Pick-ups and deliveries + urban management 623,000 200,600 790,000 traffic except shopping trips % 6% 6% 4% Shopping trips (inner, entering, outgoing) 1,403,000 236,600 1,750,000 7% % 9% 13% % Urban goods movement 13% 13% 19% Freight through traffic (harbour traffic in the 544,000 68,400 180,000 case of Marseilles) Private individuals trips (other than 13,360,000 3,020,000 10,500,000 shopping) (inner, entering, outgoing) Total 15,930,000 3,500,000 13,000,000
The environmental assessment of urban goods movement
213
The environmental model The environmental model (Figure 2) uses information provided by the output of the traffic assignment model. It allows a spatial diagnosis of the impact of various traffic links on the energy consumption, air pollution and noise. It also makes it possible to appreciate the importance of UGM concerning: (i) energy consumption and emissions of main atmospheric pollutants; (ii) concentrations of pollutants in the air, according to typical weather conditions; (iii) noise emissions and distribution.
(i)
The basic methodology of the model consists in carrying out an aggregation of the individual contributions of moving vehicles to energy consumption and polluting emissions. The emission of an average vehicle of its category (PV, LGV and HGV) is characterised by an emission factor which measures, in G / km, the quantity of pollutant generated on one kilometre. Each emission factor is defined by the nature of the pollutant (CO; CO2; NOX; HC; SO2 and PM), the category of vehicle (gasoline PV, catalysed PV and GV), the driving cycle (urban, road, motorway), the average speed of the vehicle and the slope of the link, the journey time since starting (cold / hot engine). The environmental model is carried out according to the European standards model COPERT (COmputer Program to calculate Emissions from Road Transport).
(ii) Methods of measuring the concentration are numerous (chemical, analytical, continuous methods and tele-detection/remote telemetry). Moreover it is necessary to model the meteorological conditions in so far as wind, turbulence and temperature are of the greatest importance to explain the dispersion of pollutants in the atmosphere. An Eulerian model was
214
Logistics systems for sustainable cities
used to estimate atmospheric dispersion. It resolves a specific equation including a parabolic part (concerning the diffusion) and a hyperbolic part (concerning the transportation). The input data of the dispersion calculus requires geographical data (topography, road network, pattern of settlement, etc.) and weather data. From then onwards the environmental model makes it possible to carry out quantified simulations of dispersion taking account of: (a) weather contrasting situations according to the level of stability of the atmosphere; (b) time periods: average daily traffic (ADT), morning peak hour corresponding to commercial vehicles' peak hour(MPHuGM: 11 to 12 a.m.), evening peak hour corresponding to private cars' peak hour (EPH : 5 to 6 p.m.); (c) different fleet configurations (all traffic; private vehicles (PV), UGM alone and freight through traffic (FTT). Simulations of dispersions makes it possible to evaluate the concentration of pollutants generated by road traffic in the air, as well as the contribution of goods movement to air pollution. (iii) Modelling of noise is broken up into two phases: calculation of the road noise emission and calculation of the propagation attenuation to the receiver. To complete the first calculation a level of noise called "isophone reference level" is determined. This is a theoretical noise level, measured on a vehicle moving alone during an hour under conventionally determined conditions. The isophone reference emission level is given according to the type and the speed of a vehicle, road traffic conditions, the slope of the infrastructure (uphill, downgrade and no grade). On this basis, for each acoustically homogeneous section of the network, the sound level of all vehicles of the same type is determined. Regarding noise propagation, three factors are mainly considered: geometrical divergence (characterising the energy dispersion in space according to the distance), atmospheric absorption (proportional to the distance and variable according to frequencies) and ground effect (according to the sound's frequency and respective position of the noise source and the receiver).
MAIN ENVIRONMENTAL RESULTS IN THREE FRENCH URBAN AREAS The data processing sequence presented previously was implemented in Bordeaux, Dijon and Marseilles.
Global assessment of energy consumption and emissions of pollutants Global assessments of energy consumption and emissions of pollutants are calculated according to three time periods and four vehicle fleet configurations (see section The environmental model). Table 5 presents assessments of the energy used calculated by the environmental model in Dijon and expressed in tonnes of oil equivalent per hour (toe/h). In the three urban areas studied, it appears that the strongest contribution of UGM occurs when we consider the Morning Peak Hour.
The environmental assessment of urban goods movement
215
Table 5 Global assessment of consumption in Dijon (Aria Technologies, 2001) Consumption (Toe/h) ADT (average daily traffic)
Morning Peak Hour M P H U G M ( l l t o 12 a.m.)
All traffic
10.7
12.9
29
Private vehicles
7.9
7.3
23.5
UGM (Urban Goods Movement)
2.1
4.7
4.2
FTT (Freight Through Traffic)
0.7
0.9
1.3
UGM
20%
36%
14%
UGM + FTT
26%
43%
19%
Evening Peak Hour EPH (5 to 6 p.m.)
The global assessment of emissions are calculated for six pollutants (section 3.2) and expressed in kilograms per hour. Figure 3 presents the simulated flows carried out in Bordeaux showing the importance of freight through traffic.
Figure 3 Global assessment of emissions in Bordeaux (Aria Technologies, 2000) Bordeaux city is typical: freight through traffic (FTT) takes a large place, because of its ring road or beltway that is used by many vehicles travelling between France and Spain. We can notice that FTT is essentially composed of diesel engines, generating low CO emissions and high PM and SO2 emissions. Analyses carried out on the three pilot cities show, in particular, that emissions are correlated with the number of vehicles-kilometres. Thus, emissions due to UGM are higher for the
216
Logistics systems for sustainable cities
morning peak hour which represents in Bordeaux, Dijon, and Marseilles respectively 18 %, 25 % and 19 % of the total traffic. Concerning SO2, PM and NOX, the part of FTT is always markedly higher than this part is in the total traffic. Within total UGM, heavy vehicles represent a significant part of the atmospheric pollutants emissions.
Simulation of atmospheric pollutants dispersion Within the framework of UGM environmental assessments, 24 quantitative simulations of dispersion were carried out in each of the three urban areas. These simulations are defined according to the three time sections and the four fleet configurations mentioned before as well as two weather situations (a regular situation and a detrimental one). Such simulations attempt to estimate the concentration of road traffic air pollutants just as the contribution of freight transport to air pollution. The results of these simulations describing pollutant concentration at ground level are shown in the Tables 6 and 7. Table 6 Peak concentrations calculated for a regular weather situation in Marseilles (Aria Technologies, 2000a) MARSEILLES: PEAK CONCENTRATIONS (in town centre, in |ig/m3) Regular weather situation CO HC PM SO 2 NOx CO 2 All traffic 172 17 21 1 3,005 0.6 Average Private vehicles 150 10 17 0.4 2,140 0.5 daily UGM 25 7 4 0.5 0.2 826 traffic FTT (Freight 2 (ADT) Through 0.6 0.3 0.1 0.05 178 Traffic) UGM (Urban Goods 15% 41% 27% 19% 50% 33% Movement) 15% 33% UGM + FTT 53% 20% 60% 42% hi an average situation, UGM contributes to more than 40% of the peak concentrations of NOx and PM (due to the large part of diesel vehicles) and less than 20% of CO and HC (pollutants essentially produced by gas vehicles). At MPHUGM peak, the part of UGM NOx reaches 56%, raising the total traffic peak concentration to 70 ng/m 3 , beyond the threshold of 40 ug/m3. hi Bordeaux, the 200 jxg/m3 threshold is overrun in detrimental weather conditions. Generally, the differences noticed in various cities are due to four features: city size, urban topography, network shape and weather conditions.
The environmental assessment of urban goods movement
217
Table 7 Share of several UGM indicators (%) (Aria Technologies, 2000a,b, 2001) % UGM (average daily traffic) Vehicle-km Fuel consumption
Bordeaux
Dijon
Marseilles
13
13
19
15
20
24
9 24 12 42 25 20
13 35 16 32 44 26
% Emissions CO NOx HC PM SO2 CO2
7 15 9 20 18 15
% Peak concentrations (detrimental situation) CO NOx HC PM SO2 CO2
6 17 9 20 19 17
8 23 12 36 25
20
14 46 18 33 30 28
Table 7 shows that the part of UGM external effects in Bordeaux is lower than the two other cities. It is due to the large place of freight through traffic. In Dijon which is the smaller city, there is less UGM traffic (km) but fuel consumption and PM emissions are more important due to a larger proportion of heavy goods vehicles. Peak concentrations are located in the city centres and present slight changes in comparison to average daily emissions.
Noise emissions Noise emissions are calculated on each network link according to the three time sections and the four fleet configurations noted before. Figure 4 shows the equivalent noise level (Leq) of each network section in Bordeaux (Leq measures the mean sound level, in dB(A) for a given period). Here we have noise emissions on the network of Bordeaux, concerning the UGM vehicles based on average daily traffic (ADT).
218
Logistics systems for sustainable cities
Figure 4 Noise emissions (UGM-ADT) in dB(A) in Bordeaux (Aria Technologies, 2000b / Systems Consult-Polydrom) In Bordeaux three noise level categories have been considered (below 55 dB(A), 55-65 dB(A) and over 65 dB(A)). Table 8 shows the distribution (in %) of relevant distances on the network (around 2,500 km long) according to traffic types and various time periods. Table 8 Distribution of the distances subject to different noise levels in Bordeaux (in % of the total mileage of the network) (Aria Technologies, 2000b / Systems Consult-Polydrom) < 55 dB(A) MPH U O M
All Traffic
ADT 35%
Private vehicles UGM FTT
34% 48% 47%
> 65 dB(A)
55 - 65 dB(A) EPH 60%
ADT 25%
MPH U Q M
39%
MPH U G M 62%
32%
EPH 34%
53% 39% 39%
59% 42% 38%
72% 45% 45%
12% 14% 14%
13% 21% 15%
16% 11% 12%
ADT
5%
EPH 6%
28% 38% 47%
12% 45% 43%
The estimates above show that approximately a third noise levels higher than 65 dB(A) in the peak hour average daily traffic. Concerning UGM, the part of dB(A) is definitely more significant for MPH UGM than
of the Bordeaux network is exposed to (MPH or EPH), against a quarter with the network exposed to more than 65 for other time periods.
CONCLUSION The comparison of the various estimated traffic segments shows similar proportions from one city to another. Moreover there is no inconsistency between our results and those of specific studies undertaken by various local planners, hi the same way, the modelled emissions and dispersions levels we obtained are consistent with the observations done by air quality check networks. However, the various results are subject to risks, for three main reasons: a strong heterogeneity of information sources, a lack of knowledge for certain traffic segments and complexity within the processing chains.
The environmental assessment of urban goods movement
219
The data heterogeneity is essentially due to: a spatial dimension (definition of areas studied and zoning systems), a temporal dimension (e.g. shopping trips are measured on average daily traffic while UGM is measured weekly), different observation levels (e.g. goods movement, vehicle flows and passenger trips). Flows generated by building sites, road works and work areas are poorly understood and there is only limited information about the spatial distribution of these traffic flows. On one hand there is an important complexity in data-gathering, on the other hand this complexity is even more significant when we consider the question of modelling. In particular, if no drivers' routes surveys are available, it is necessary to implement a traffic generation model associated with deliveries, like the FRETURB model (Routhier et al, 2001) used in France.
ACKNOWLEDGEMENTS The authors would like to thank Aria Technologies (Mr A. Albergel and Mrs A. Fresneau) for the development and the implementation of the environmental model and Systems Consult (Mr C. de Rham) for the development and the implementation of the assignment and noise model. This research was funded from ADEME (French environmental agency) and EDF (French electricity board).
REFERENCES Ambrosini, C. and Routhier, J.L. (2003). Objectives, methods and results of surveys carried out in the field of urban freight transport: an international comparison. Transport Reviews (forthcoming). Ana Technologies (2001). Bilan environnemental du transport de marchandises en ville. Agglomeration de Dijon. Final Report for ADEME and EDF. Aria Technologies (2000a). Bilan environnemental du transport de marchandises en ville. Agglomeration de Marseille- Final Report for ADEME and EDF. Aria Technologies (2000b). Bilan environnemental du transport de marchandises en ville. Communaute Urbaine de Bordeaux. Final Report for ADEME and EDF. Boerkamps, J. and Binsbergen, A. van (1999). GoodTrip — A new approach for modelling and evaluating urban goods distribution. In: E. Taniguchi and R.G. Thompson (eds./ City Logistics I, Institute of Systems Science Research, Kyoto, 175-186. Certu (1998). L 'enquete menages deplacements methode standard- CERTU, Lyon, France. COST 319 (1999). Estimation of pollutant emissions from transport. Final Report of the Action, Brussels. Gallez, C. and Hivert, L. (1998). BEED: mode d'emploi. Synthese methodologique pour les etudes Budget Energie Environnement des Deplacements, INRETS Final Report. Ma, L. (1999). Modelling for traffic-generated environmental pollution analysis. In: E. Taniguchi and R.G. Thompson (eds.). City Logistics I, Institute of Systems Science Research, Kyoto, 133-146. Routhier, J.L., Segalou, E., Albergel, A., and Rham C. de (2002). Mise en place d'une methodologie pour la realisation des bilans environnementaux du transport de marchandises en ville. Research Report for ADEME. Laboratoire d'Economie des Transports, Lyon, France.
220 Logistics systems for sustainable cities Routhier, J.L., Segalou, E. and Durand, S. (2001). Mesurer I'impact du transport de marchandises en ville. Le modele de simulation FRETURB (version 1). ADEME-DRAST, Paris, France. Routhier, J.L., Patier, D. and Ambrosini, C. (1996-99). Transport de marchandises en ville: resultats des enquetes quantitatives de Bordeaux, Dijon et Marseille, Final Reports for DRAST. Laboratoire d'Economie des Transports, Lyon, France. SES (1999). L'utilisation des vehicules utilitaires legers en 1996, Ministere de l'Equipement, du Transport et du Logement. Valentini, M.P., Lacquanti, P. and Valenti, G. (2001). Methodology and results of a study on logistic schemes in Siena. In: E. Taniguchi and R.G. Thompson (eds.). City Logistics II, Institute of Systems Science Research, Kyoto, 185-199.
16 ROUTE CHOICE AND THE IMPACT OF 'LOGISTIC ROUTES'
Dr. Jaap Vleugel, Technical University Delft, OTB Research Institute for the Built Environment, Delft, the Netherlands Dr.-Ing. Milan Janic, Technical University Delft, OTB Research Institute for the Built Environment, Delft, the Netherlands
ABSTRACT Urban goods transport has emerged as a complex issue in increasingly congested urban areas. Vehicles are often trapped in congested streets where drivers also find it difficult to park and load or unload. Time-pressure becomes high, especially when a delivery vehicle is involved in a multidrop trip. In addition, regulation may restrict vehicle dimensions and weight, access to an area, stay time in the area and environmental burdens in terms of air pollution and noise, which additionally hinders using a vehicle in the most efficient way. In some cases route choice is not free. This paper investigates the route choice of truck and van drivers in selected Dutch cities, when they are affected by supply (infrastructure, regulation, etc.) and demand constraints. Such situations result in suboptimal route choice. Data from interviews and questionnaires among drivers, receivers of goods and persons living in the area where goods are loaded or unloaded are used. The use of real data instead of models is a relatively new approach, which sheds new light on the subject of urban goods distribution.
INTRODUCTION
Goods city distribution Recently distributors of goods in cities are faced with growing congestion and diminished accessibility ', caused by developments in logistics, transport and in local policy. In logistics, order-driven production (OECD, 1992; ECASC, 2000) means more, but smaller orders, more 1 For example, in the past 20-30 years average speed of goods city distribution went down from 25 to 10 kilometers per hour in The Netherlands (Rijssenbrij, 2002).
222 Logistics systems for sustainable cities frequent deliveries, small or no stocks and shorter lead times. Transport has to adapt itself to these demands. With smaller average order sizes, trucks make multidrop (round) trips delivering goods at many locations within the same city or region. The local situation (narrow or one way streets, lack of parking space or stocking depot) may demand use of city distribution vehicles. For the same volume more vehicles are needed and they also tend to have a lower load factor (on average) 2 and thus less consolidation. This may increase delivery costs, stress on the environment and liveability3. Stress on drivers also increases, which may cause more road accidents. Reduced accessibility may also have another impact, namely that some receivers may decide to leave the inner city in favour of suburban or out-of-town business locations, which are more easily accessible by freight and passenger vehicles. This may also reverse modal split in passenger transport 4 and increase suburban congestion.
Policy Local governments usually want to reduce the externalities caused by distributing goods in cities. Dutch research has shown that the share of freight vehicles in urban traffic ranges from 1% for local roads to 18% for through roads. Part of it concerns transport of waste, road building materials and its removal. Delivery is mainly done by light vehicles (Bouman et al., 1990). Goods transport in urban areas is governed by policy instruments mainly developed for passenger cars, such as one way streets and dedicated pedestrian areas. Special instruments for goods vehicles include access restrictions, most notably time windows, physical restrictions (size and axle weight) and loading restrictions (location, number and size of parking places or stops). Their impact on the distribution of goods in cities may differ, depending on the way these instruments are used. A delivery-oriented policy for a shopping centre is likely to have a different impact on delivery than a liveability-oriented one. Policy may also differ between cities or towns, even if they are located next to one another. Lack of harmonization is a major problem for truck drivers involved in regional delivery (Dutch Platform on City Distribution; PSD, 2001).
THE SYSTEM AND THE PROBLEM Background Urban goods transport is the last (or first in case of reverse logistics) leg of the chain from suppliers to receivers. Together with wholesalers, logistic planners, vehicle drivers and policymakers are responsible for decisions regarding or influencing: vehicle choice (fleet management); trip planning (frequency of shipment, duration of trip, etc.); Frequently a mismatch exists between the internal dimensions of a vehicle and those of a certain load unit, e.g., a small container. This may reduce the load factor significantly (Rijssenbrij, 2002). The impact of goods city distribution vehicles on liveability and the environment cannot be assessed easily. The assumption that larger vehicles have a lower overall environmental 'footprint' than smaller ones is not confirmed in the literature. 4 This happened for instance in Amsterdam after the opening of the southern business area. The innercity lost at least 40% of its jobs. 3
Route choice and the impact of 'logistic routes'
223
route planning (use of road infrastructure); the decisions of other actors in or outside the supply chain. This paper focuses on the determinants of route choice by drivers of delivery vehicles in urban areas. In Figure 1 a purchase order from a receiver is received by the nearest distribution centre (DC). If the goods are available5, logistic planning is used to choose an optimal transport operation. Then transport from the warehouse to the receiver takes place. In most cases the DC is located outside an urban area. This means that a regional trip towards the urban area is needed. The final part of this trip, the urban leg, is dealt with in this paper.
Figure 1 Relations between orders (O), logistic planning (L), transport (T) and local policy (L) Previous research Research into route choice is extensive. The standard approach is quantitative and uses mathematical optimisation techniques (e.g., Taniguchi et al., 2001). A simplified model of the actual road network is fed into a computer. Algorithms will then calculate the optimal routes. Embedded in these algorithms is the assumption of rational behaviour of drivers. Route choice is improved by better information (systems). A government could use the same information systems for 'real-time' monitoring of drivers. This enables flexible routing and more efficient use of roads (OECD, 1992). The actual driving experience is much different. Uncertainty is a key problem. First, the road network is much more complex, so there may be more alternative routes. Second, congestion may suddenly emerge, increasing travel time. Space for parking may not be available when needed. Most decisions are not made rationally (Tversky and Kahneman, 1986). Human capability to process information generated simultaneously by different sources is limited. People tend to use rules of thumb experience instead. Finally, real time traffic information is not available for delivery streets or areas. How can these constraints be dealt with? Probability-based techniques, such as fuzzy-set theory, may help to some extent (Taniguchi, et al., 2001). Most importantly however is that more information about how drivers actually choose their route and why they do so is required. This could then be used to validate the available models and to design new models. If the goods are not available in the nearest DC, they have to be transported from the next higher one.
224
Logistics systems for sustainable cities
Objectives of the paper This research is based on the behaviour of real freight drivers by statistically analysing questionnaire data from shopping streets in four Dutch cities (PSD et al., 2002a and 2002b). Local policy-makers supplied data for the background study and were also active in an expert panel. The following research questions are investigated: 1) How do freight drivers choose their routes? 2) Which constraints on route choice do drivers mention and how do they deal with them? 3) What is the impact of local policy on route choice? The paper introduces the background study, the questionnaires and the methodology used to analyse the data. An analysis of the data-set is presented concentrating on route choice is then presented. Finally, the main conclusions and recommendations are given.
METHODOLOGY The background study Lack of knowledge, both scientific and practical, prevents local decision-makers from developing specific policies relating to goods transport. They tend to respond to signals they receive from society, without fully understanding what is going on. With more knowledge, a balance between the demands of goods delivery and social needs (eg. noise level and air pollution) can be achieved. A study commissioned by Connekt (Knowledge Centre for Transport and Traffic), Delft, The Netherlands dealt with the following research question (PSD et al., 2002a, p. 8): "Can dataacquisition for city distribution be undertaken in a structured and consistent way with the purpose to use the data for an explanation of goods distribution processes in cities, both in a quantitative and a qualitative sense?" The data-collection should allow design of a so-called goods delivery profile of a city area (shopping street or area etc.) which evaluates - Economic vitality and attractiveness: the visual quality of the area, type of economic activities and their development over time. Traffic safety: the share of delivery traffic in traffic accidents and the type of damage caused by freight vehicles. Liveability: experienced noise and air quality. - Accessibility: availability of delivery time windows and physical blocks, roads used etc. Quality of delivery: subjective judgements by the actors involved of strong and weak elements, kind of solutions to improve these elements. A few 'good' in-depth studies were used to test the method to develop the goods delivery profile. Such studies have more than local importance, because they can contribute to the understanding of goods city distribution in Dutch cities. Either the standardized method or its results could then be transferred 6 to the cities in The Netherlands, which would remove the need to carry out such a study for each particular city. By fine-tuning the 'standardised' knowledge to the local situation, policy-making could be improved. In environmental economics, this technique is called benefit transfer (see Bergland et al., 1995).
Route choice and the impact of 'logistic routes'
225
The background study consists of two stages. In stage A, the method to collect and analyse the data was developed. Collection and first analysis of the data took place between October 2001 and May 2002. In stage B, more data will be collected, in other shopping centres in The Netherlands. For complementary information questionnaires will also be sent to residents. The main purpose of stage B is to develop generalised indicators for goods distribution in cities, which could be used to show the impact of policy measures in a particular city on goods distribution, without the need for in-depth analysis of the actual distribution situation in the same city. In this paper, data from stage A are used.
Data-collection Introduction Data-collection took place in four shopping areas in the dutch cities Amsterdam, Alphen aan den Rijn, Apeldoorn and Rotterdam; two large and two middle-sized cities. Questionnaires were developed for receivers, transport companies, truck drivers and shippers (logistic managers). Additional information was received from the governments involved in the project. Lack of time and an insufficient budget prevented counts of incoming and outgoing freight vehicles. Choice of respondents For practical and budgetary reasons, the sample size per actor group aimed for was sixty. In case of receivers this was achieved and they were also evenly distributed among the following branches: 1. Supermarkets; 2. Daily retail; 3. Fashion; 4. Equipment/furniture; 5. Other retail; 6. Hotel, catering and entertainment; 7. Services and institutions; 8. Other. In the case of shippers and transport companies the group was less than sixty, this is why the whole group was chosen. Drivers were picked on the spot, because further stratification into subgroups is difficult and in fact unnecessary for the analysis. Respondents were chosen as follows. The Chamber of Commerce delivered the base data (address, branch of industry, etc.) of the receivers. Drivers were directly contacted while loading or unloading. The contact data of shippers and transport companies was acquired by asking receivers to mention the name of their shippers and transport companies. Shippers were asked to supply the names of the transport companies involved. Drivers were asked to mention the firm they worked for. Table 1 shows the group and sample sizes for each actor group. Table 1 Groups Total Receivers All 2899 Personally visited 237 Shippers All 422 Phoned 243 Respondents 124 Transport companies All 310 Phoned 171 Respondents 110
and sample size for each actor group Amsterdam Apeldoorn Alphen 1312 863 386 56 61 59
Schiedam 338 61
36
29
30
29
26
25
33
26
226
Logistics systems for sustainable cities
Freight drivers All Questionnaires Respondents
315 315
79
79
78
79
It was not so easy to get information from shippers and transport companies. In various cases, drivers and receivers were unable to supply their names and addresses. Once contacted, many of them were not very willing to provide information about their logistic decision-making processes. This experience casts doubt on whether personal interviews with logistic managers provide more information. It costs much more time (and money) to visit them, though.
DATA ANALYSIS Delivery profiles Route choice is related to all major aspects of the delivery profile. A short presentation of such a profile will be presented initially. The sample size does not warrant a direct calculation of the delivered freight volumes, number of trips, etc. in the four cities. The four cities were taken together and regarded as one sample (PSD, 2002a) 7. The average number of deliveries per week is calculated and multiplied by the total number of firms (2899). A similar calculation can be made for the volumes 8 involved. Table 2 gives an overview of the results of this 'upsampling' for all four cities. Table 2 Freight traffic in the four cities Amsterdam Apeldoorn Schiedam Average Alphen 1312 863 341 Number of firms 386 6720 m J 6662 m3 2209 m J Freight volume/week 2348 m J 4485 m J 8779 4881 2818 Deliveries/week 3078 4889 6.2 4.7 1.8 Average number of stops 3.9 4.15 per trip 1417 1417 1297 1566 Number of delivery trips 789 /week 11% 39% 22% 58% 26% Pass car or van (cat. 0) " 32% 29% 10% 27% Light truck (cat. 1 ) / ( 30% Truck (cat. 2 ) " 27% 28% 42% 52% 49% 7% 1% 0% 3% Truck + trailer (cat. 3) 4 ) 5% Notes: 1) Cat. 0 = < 3.5 tons ; 2) Cat. 1 = 3.5-7.5 tons; 3) Cat 2. = 7.5-18 tons; 4) Cat. 3. = > 18 tons (Source: PSD, 2002a) The delivery information is part of the economic information in a delivery profile. Table 3 gives an example for the city of Alphen.
7
This is a proxy, assuming that the four cities are similar, which they are not. Or the average number of employees or square meters of selling space. These data were not (sufficiently) available for all cities, however. 8
Route choice and the impact of 'logistic routes'
227
Route choice Route choice is determined by the match between constraints set by receivers (order size, delivery moment etc.), shippers (opening hours of DCs) and logistic planners/fleet managers (loading factor, cost) and on-street conditions. Ultimately it is the driver who chooses the route which enables him to visit all the receivers during one trip before returning to its home base. Table 3 Delivery profile for Alphen aan den Rijn (summary) (Source: PSD et al., 2002b:12) Economic vitality • The study area contains 386 firms, receiving 2348 m3 goods, divided and attractiveness among 3078 deliveries Traffic safety • A truck of category 2 (7,5 - 18 tons) is used for 52% of all deliveries • Freight vehicles are involved in 33% of all accidents in the city centre and in 6% of accidents outside the city centre Liveability • Insufficient data to draw conclusions at the moment Accessibility • Accessibility of the city centre is regarded as reasonable to good • Receivers are hardly hindered by time windows Quality of delivery • The general feeling is that the delivery situation is good • The bridge crossing the Old Rhine is an important barrier for delivery traffic • The quality of the pavement is mentioned as a weak point To elaborate on route choice the following questions were answered: - What are the main access routes used by delivery traffic? - Which supply factors restrict route choice? - How many drivers follow regular (fixed) routes? What is the influence of demand and supply constraints on regularity of route choice? Main access routes In and outgoing traffic uses the road links between highways and city edges. Individual drivers were asked to draw main access routes in a map 9 and to express their opinion about the quality of accessibility. The answers were used to derive main access routes for each city and the spread of traffic over these access routes. According to drivers external accessibility is good in all cities. In a larger city it takes more time to enter and leave, just because of the travel distances and traffic volumes. Table 4 gives an overview of local accessibility in the four cities. This shows that drivers were more positive about the accessibility of shops than receivers, shippers and transport companies. This is consistent with other research saying that the perception of a situation is more negative among those who never experienced it compared to those who did. Policy-makers, while promoting their policy, are usually the most positive about accessibility. Accessibility in the city itself is determined by many factors, some of which are the same as for external traffic. Freight drivers usually try to evade barriers by changing route or time of delivery, but this assumes that alternative routes are available.
Only in Amsterdam time prevented most drivers to do that correctly. Linking the sketches with real streets turned out to be difficult. Drivers either lack the time or knowledge to draw useable maps. This is why our effort concentrated on determining main access routes between the highway and the streets leading to the innercity.
228 Logistics systems for sustainable cities Only Apeldoorn offers alternative routes of sufficient quality, which, as various drivers mentioned, can be used if the most preferred road is blocked. The small number of passenger cars makes delivery a lot easier. In Alphen a major bridge cannot be used by delivery vehicles. This leads to detours. Road maintenance and other delivery traffic (eg. moving vans) are also mentioned as causing road blocks (hence a reason for a detour). In cities with time windows (three out of four), the time factor influences route choice. Constraints on route choice The following determinants and barriers for route choice could be distinguished Quality of the infrastructure; Quantity of the infrastructure, including alternative routes; Available parking options (and distance to shops); Conflicts with passenger cars; Road maintenance; Other delivery traffic; Time windows.
10
:
Regularity of route choice Delivery route planning involves concentrating orders for a specific area, thereby effectively optimising trip time and transport operations. Choosing a different route is often timeconsuming and thus is only done if necessary. Following this assumption, drivers will likely rely on past experience and repeatedly use the same route. Figure 2 shows that most drivers, 70-80%, follow regular routes. Amsterdam and Schiedam have a lower score. Figure 3 indicates that in some specific branches of industry fixed routes are more common than in others. Route choice receives alternative routes. In Amsterdam, such routes are hardly or not available at all. One way streets and lack of available parking places lead to many unsuccessful searches for alternative parking lots. In Schiedam, narrow streets with many parked cars and other freight vehicles either moving or unloading and closed streets explain the proliferation of small freight vehicles, including many passenger cars. Some truck drivers suggested city distribution centres as a way to improve the delivery situation in this city. Determinants of regularity of route choice If drivers do not follow the same route again and again, order behaviour may be a prime cause. Demand may fluctuate strongly on a day-to-day basis 12. One may assume that in industries where customer orders determine production route choice is more irregular than in industries which are more supply-oriented. Manufacturing is an extreme example. Then there is the impact of order changes. Orders can be eliminated, changed or added during the day, which makes it likely that delivery routes are changed (Bradford, 1999).
10 Most of the 'scores' on these variables were put in a (public) database. Maps were made showing main shopping centre(s) and pedestrian area(s), parking places for trucks, one way streets, vehicle restrictions etc.
" Statistical analysis to quantify the influence of each supply and demand constraint on route choice did not lead to valid results. This also holds for possible interrelations between these factors. It becomes apparent, that a conclusion at a more general level about good accessibility may ignore specific local problems. 12 Following the ECR concept. In the optimal case, larger trucks should be used for let's say 80% of demand volume and small vans for the remaining part. In practice fleets are likely to consist of average size delivery vehicles. This may lead to fluctuating loading factors, depending on the 'urgency of delivery'.
Route choice and the impact of 'logistic routes'
229
Another reason for irregular routes may be that the drivers do not have many deliveries in the city (no round trip). The city is probably part of an interregional trip 13 A high frequency of delivery (per day) does not (always) have to be associated with irregular routes, as the case of supermarkets shows, however. Supply conditions, most notably road maintenance works, can have a strong impact on route choice, especially if alternative routes are not available. Amsterdam is infamous for this. If a long main shopping street is completely reconstructed during many years, and 'step-wise' certain access roads are closed, drivers have to change access routes accordingly.
Figure 2 Regularity of route choice (Source: PSD et al., 2002b)
The available data does not permit setting-up an O-D matrix.
to us
Table 4 Main barriers reducing internal accessibility and influencing route choice (Source: PSD et al., 2002b) Time windows (allowed to drive and un-load) % hindered by time windows Driving by night allowed Delivery outside time windows allowed Regulated vehicle dimensions Physical barriers Main vehicle type used Infrastructure Number of (un)loading bays (un)loading on the street dimensions of bays location of bays vis-a-vis main roads narrow streets or cays damage to houses etc. One way streets/routing
o
Alphen (Mo-Fr) 04-12.00 (Sat) 04-08.00 21% No (ban, few exemptions) No No No Large
Apeldoorn 3> Amsterdam Most of the area 09-19.00 (Mo-Sa) 06-11.00 (depends on street) 55% 39% 2) Yes Yes No Yes No No Yes (bus sluice in city centre) No Large Mainly pass car
Sufficient No Sufficient
Insufficient Yes Sufficient
Sufficient No Sufficient
Sufficient No, but on the pavement Sufficient
Good No No Yes
Bad Yes Yes Yes
Good No No Yes
No (except Ten Kate market) 39%, 41%, 52%, 40%
Yes" 57%, 62%, 59%, 58%
Good Yes Yes Yes (route signs are mentioned as inadequate) Partially 43%, 52%, 43%, 50%
Problematic
Good
Good
Semi-closed pedestrian area (gates) Yes" Accessibility is good to very good 72%, 56%, 38%, 67% according to drivers, receivers, shippers and transport companies Rating internal accessibility (overall) Good
Schiedam (Mo-Fr) 07-11.00+ 18-19.00 56% Yes No Yes (except for height) Yes (narrow streets) Pass car + van
Notes: 1) If on time in and out, the gates are not a problem for delivery traffic; 2) Presumably hindrance from time windows in other parts of the city (during a round trip); 3) 70% of the receivers found no reduction in sales due to the closure of streets, while 61.5% mentioned no additional delivery problems.
I
I ! S' o o-
I
Route choice and the impact of 'logistic routes'
231
Table 5 Time needed to reach, to stop and stay in the area
Time in minutes
Alphen
Amsterdam Apeldoorn
Average time to 16 25 reach the area 10 Mode of time to 10 reach the area Average time in the 43 91 area (stay time) 15 Mode stay time 60 Average time to 27 21 (un)load (stop time) Mode stop time 15 10 (Mode, is the most frequent response)
Schiedam
18
6
Average of the 4 areas 16
15
5
10
52
30
54
30 23
30 22
30 23
15
10
10
Accessibility of an area may be assessed in terms of driving, stay and stop time in the given area. Stay time may be divided into time to drive, time to load or un-load and time to wait. Stop time is the time needed to load or un-load, so stop time is part of the stay time. Table 5 gives an interesting overview for the four cities. Schiedam has a low average stay and stop time. Table 2 tells that Schiedam has the smallest number of stops per trip, which conforms with the frequent use of very small vehicles. The average stop time is not shorter than in the other cities. So, instead of unloading larger units from a larger vehicle, in this case many small boxes are unloaded from a small vehicle. In Amsterdam, the low average stop time is influenced by the bad parking situation (on street, easy to get a fine), hence the average volume delivered is also relatively small. Stay time is high due to the many stops per trip and the traffic situation.
232
Logistics systems for sustainable cities
Impact of policy on route choice Background Local policy may restrict access to a certain area (eg. closed routes, one way streets) or concentrate access (eg. preferred routes). Preferred or obliged routes are common for freight vehicles carrying dangerous goods or goods of non-regular size; delivery traffic not meeting specific criteria (eg. weight per axle, width and size); public transport vehicles (eg. bus lanes, tramways and metro tunnels); fast motorized traffic (eg. bypasses or ring routes for through traffic); carpool vehicles. Obliged routes imply 'routing' of freight and other traffic. A 'circulation plan' may discriminate between through and non-though traffic. Routing aims to protect liveability, to increase traffic safety and to prevent damage to roads or buildings. For goods delivery traffic routing can be used together with time windows. Improvements in terms of liveability in certain streets may lead to inaccessibility in other streets, shifting problems from one place to another. From the data it becomes apparent that shippers and transport companies respond to poor accessibility by using smaller vehicles (vans or even passenger cars instead of trucks) in the area involved. Schiedam and Amsterdam are typical examples. Policy in the four cities Routing of goods delivery traffic in the four cities looks is described here. Alphen has a semiclosed city centre with large shopping area. There is one fixed barrier (the bridge crossing the river Oude Rijn). There are time windows with little exceptions for goods delivery vehicles. A masterplan to upgrade the city centre has been launched. It includes routing to improve the delivery situation. The direction of future policy may be less restrictive and more facilitating. The situation in Amsterdam Old West is different. There is no comprehensive plan for delivery traffic, there are one-way streets and no routing. Parking in bays in front of shops should be possible, yet the shortage of bays leads to a lot of parking violations by delivery vehicles. Future policy direction is likely to be more restrictive relating to vehicle access. In Apeldoorn the situation is similar to Alphen with a semi-closed city centre with shopping area. There are one way routes, but there is no routing. Future policy may become more restrictive for large vehicles. Schiedam has a main road around the city centre that helps to keep through traffic out of the city centre and to guide delivery traffic to their locations. There are many roads closed for freight traffic and one way streets and also flexible gates, but there is no routing. Future policy may become more facilitating, in the sense that route signs will be improved and enforcement of parking regulation will become tighter. The absence of routing in most cities is also seen in other European cities (OECD, 2003). hi various cases the number of trucks, their dimensions or emissions are regulated by permits. Only in a very few cases trucks are actually banned from a given road or area.
CONCLUSIONS AND RECOMMENDATIONS This paper has shed light on route choice by drivers of freight vehicles. The main routes chosen by drivers to enter and leave cities were described. In case of internal traffic major bottlenecks
Route choice and the impact of 'logistic routes'
233
have been identified. Next to physical constraints, local policy measures in many cases worsen the delivery situation. Drivers are forced to stay longer in the city than in a situation with less barriers. They are in most cases not free to choose their route, and there's no certainty that the routes chosen are the most efficient or the least detrimental in terms of liveability and safety. This is because a traffic ban does not really discriminate between through traffic and local traffic. Policy should find ways to make this distinction. The gate system, as used in three of the four cities is a major attempt to make this distinction. In practice, it has benefits in terms of liveability and shopping 'climate', but has some weaknesses, with respect to inflexibility (in time), vehicle size and illegal use by shop owners. Drivers' choice of route is also constrained by demands from receivers, shippers and their own logistic planners. It was not possible to get more information about their logistic strategies and decision-making. Shippers and transport companies were not very willing to fill in the questionnaires, while receivers were not asked why they were delivered in the way the questionnaires uncovered. Personal interviews with these decision-makers could shed more light on this. The background study was based on relatively small datasets, which reduces the relevance of the data for quantitative purposes. More study is needed, also in quantitative terms, to make the analysis more generally applicable. The study in stage B will be used to check and explain the findings of stage A and enlarge the dataset.
REFERENCES Bouman, P.A., Kluit, P.J.L., Schoemaker, Th.J.H. and J. v.d. Waard (1990). Goederenvervoer en leefmilieu. Inventarisatie van emissies en verstoring door goederenvervoer, Faculty of Civil Engineering, Dpt. Traffic, TU Delft. Bradford, B. (1999). The route delivery problem, White paper, IBM Corporation. European Council of Applied Sciences and Engineering (ECASC) (2000). Freight logistics and transport systems in Europe, Euro-CASE "Freight" Steering Group, Paris. Kahneman, D. and Tversky, A. (1996). On the reality of cognitive illusions, in: Psychological Review, Vol. 103, pp. 582-591. McMahon, D. (2000). Death of the sales force, ECR model may place emphasis on service and technical support, in: The Graziado Business report, Spring 2000, The Graziado School of Business, Culver, http://gbr.pepperdine.edu/002/sales.pdf. OECD (1992). Advanced logistics and road freight transport, Paris. OECD (2003). Delivering the goods, 21st century challenges to urban goods transportPSD (2001). Duurzame stedelijke distribute, Werkboek, Den Haag. PSD, TLN, Trail, DHV, Municipalities of Alphen a/d Rijn, Amsterdam, Apeldoorn and Rotterdam (2002a). Dataverzameling stedelijke distributie, Deel 1: Methodologie bevoorradingsprofiel, Den Haag/Zoetermeer/Delft/Amersfoort. PSD, TLN, Trail, DHV, Municipalities of Alphen a/d Rijn, Amsterdam, Apeldoorn and Rotterdam (2002b). Dataverzameling stedelijke distributie, Deel 2: Bevoorradingsprofielen en onderliggende kenmerken, Den Haag/Zoetermeer/Delft/Amersfoort. Rijssenbrij, J.C. (2002). Stedelijke distributie in de toekomst, Standaardisatie en aangepaste voertuigconcepten, in: Stedelijke distributie in de retailketen, Handout bij eindpresentatie op 18 September 2002, Connekt, Delft, pp. 7-17. Taniguchi, E., Thompson, R.G., Yamada, T., and R. van Duin (2001). City logistics, network modelling and intelligent transport systems, Pergamon, Oxford.
This page is intentionally left blank
17
EMPIRICAL ANALYSIS ON HAZARDOUS MATERIAL TRANSPORTATION USING ROAD TRAFFIC CENSUS AND ACCIDENT DATA Tatsuhide Ito, Docon Co. Ltd., Sapporo, Japan Makoto Hayano, Docon Co. Ltd., Sapporo, Japan Toshiyuki Naito, Docon Co. Ltd., Sapporo, Japan Yasuo Asakura, Graduate School of Science and Technology, Kobe University, Kobe, Japan EijiHato, Department of Civil and Engineering, Ehime University, Matsuyama, Japan Tadayuki Wada, Hokkaido Regional Development Bureau, Sapporo, Japan
ABSTRACT This study aims to identify the characteristics of hazardous-load transportation in Hokkaido, based on which a scheme for new road networks to reduce the risks associated with hazardous-load vehicles will be developed for the future. The study presents some new proposals, after analysing the current situation regarding hazardous-load transportation, examining its problems and estimating accident occurrence risk rates, with a focus on two strategies to reduce the risk of accidents involving hazardous-load vehicles, choosing less populated areas as optimum routes for the transportation of hazardous materials, and completing transportation operations in a short time by choosing the shortest route. Procedures taken are as follows. First, a general formula is proposed to model the risk incurred during transportation of hazardous materials. The effects induced by hazardous materials depend upon the distance, and the number of exposure victims is given as an expected value. The risk evaluation model for transporting hazardous-loads can be expressed as the product of the expected number of exposure victims and the total travel distance. Second, the risk evaluation model is applied to the road network in the Nanyo area of Shikoku. As a method to minimise risk, two possible optimum routes, the shortest-distance route and the least-populated route, are proposed and analysed with the risk evaluation model. Optimum minimum-risk routes for hazardous-load transportation in the Nanyo network differ from OD pair to OD pair if the total travel distance is less than 60km. When the total travel distance is between 60km and 120km, the selection of the least-populated route produces a lower risk. On the other hand, the level of risk posed by the shortest-distance route becomes lower (than that of the least-populated route) when the total travel distance becomes 120km or longer. Therefore, the optimum route to minimise risk depends upon the
236
Logistics systems for sustainable cities
nature of the OD pairs.
INTRODUCTION Japan's logistics industry has played a vital role in supporting the national economy as well as assisting citizens' lifestyles and industry. This road freight traffic, however, includes a certain proportion of hazardous-load vehicles, which are capable of causing extremely detrimental effects to surrounding natural and social environments if they are involved in traffic accidents. Various studies have been carried out on the reliability of road networks to sustain efficient logistic operations. These studies include evaluation of such factors as origin-destination (OD) connections and time reliability. Yet, few studies have been conducted in relation to risk assessments and reliability analyses of risk-inherent traffic or hazardous material transportation. Therefore, comprehensive evaluation of road-network reliability, which also reflects risks entailed by hazardous-load vehicles, is a critical issue to be addressed. This study aims to identify the characteristics of hazardous-load transportation in Hokkaido, based on which a scheme for new road networks to reduce the risks associated with hazardous-load vehicles will be developed for the future. More specifically, this study will present some new proposals, after analysing the current situation regarding hazardous-load transportation, examining its problems and estimating accident occurrence risk rates, with a focus on two strategies to reduce the risk of accidents involving hazardous-load vehicles, choosing less populated areas as optimum routes for the transportation of hazardous materials, and completing transportation operations in a short time by choosing the shortest route.
INTERREGIONAL FREIGHT MOVEMENTS IN HOKKAIDO Freight movements in the Sapporo area (both as an origin and a destination) account for 27-29% of the total volume of interregional freight transportation in Hokkaido, thus making Sapporo Hokkaido's biggest logistics base. The annual volume of freight transportation in Hokkaido totals Table 1 Total freight transportation in Hokkaido Sapporo
Asahikaw a
Hakodate
Muroran
Kushiro
Obihiro
(Unit : thousand tons] Total in Kitami Hokkaido
Volume 69,250 70,666 76,224 593,714 158,208 40,293 88,814 90,259 exported Share (11.7) (12.8) (100.0) (26.6) (15.2) (6.8) (15.0) (11.9) (%) Volume 71,345 71,882 77,025 173,966 92,730 42,496 73,515 602,959 imported Share (11.8) (12.8) (100.0) (28.9) (15.4) (7.0) (12.2) (11.9) (%) Source) "1997 Freight Transportation Survey" by the Transport Policy Bureau, the Ministry of Transport
Analysis on hazardous material transportation
237
567 million tons, among which transportation within the Sapporo area is the highest with an annual volume of 142 million tons, followed by 84 million tons per year within the Ashahikawa area and 73 million tons per year within the Kitami area. As for interregional freight, shipments between the Muroran and Sapporo areas are the highest within Hokkaido with an annual volume of 18.8 million tons. This is understandable as the Muroran area is serviced by both Tomakomai Port and Muroran Port. Regarding the distribution of freight shipments from Sapporo, Muroran receives the largest proportion with an annual volume of 5.5 million tons, which is followed by the Asahikawa region receiving an annual 2.9 million tons. In terms of the modal share, 98.1% of the total freight transportation in Hokkaido depends on road transportation.
Figure 1 Interregional freight transportation in Hokkaido
238
Logistics systems for sustainable cities
CREATION OF NEW EVALUATION INDEXES FOR ROAD NETWORKS
Evaluation indexes for road networks Table 2 shows aspects, indexes and methods generally used in evaluating road networks, in which evaluation is carried out using various evaluation indexes related to transportation functionality, environment, lifestyle and land use. Some aspects such as transportation functionality have already established simulation analyses while some others include unusual but important indexes such as disaster prevention and risk reduction. Table 2 Aspects and indexes of road network evaluation Aspects Transportation functionality Travel time Travel cost Traffic accidents Supply-demand balance Improved mobility Comfort Environment Air pollution, noise Ecosystem Lifestyle Social exchange Public service Disaster prevention Risk reduction Land use Industrial vitalisation improved road network Land usage
Methods
Indexes Cost of lost time Cost of trip Cost of accidents Degree of congestion Travel vehicle-km, travel vehicle-hour, etc. Reduced fatigue
Traffic Traffic Traffic Traffic Traffic CVM
assignment assignment assignment assignment assignment
simulation simulation simulation simulation simulation
Emission of NOX and CO2, noise level Habitat of rare species, soil and water quality level
Traffic assignment simulation Environmental assessment
Travel time between major nodes Covered area of a major node No. of disaster (risk) inspection points, etc. Human and material damage in an accident involving hazardous-load transportation
Time space estimation Time space estimation Redundancy analysis ?
No. of newly established enterprises and projects Density of road network Change in land use
Statistical analysis Density of road network Land use analysis
Selection of evaluation factors In this study, a new method of evaluation was researched which was based not only on typical evaluation factors, but also on the "risks induced by hazardous-load transportation." In addition, consideration was given to the resulting social implications as well as the project's significance as a research theme. The social importance associated with road networks is attested by the following facts: a certain proportion of vehicles transporting hazardous substances such as inflammable or poisonous materials are included in road traffic; they are capable of causing extremely detrimental effects to surrounding natural and social environments once involved in an accident; thus they are a potential factor to be considered in the evaluation of a road network improvement project. Hazardous-load transportation also provides a new perspective as a research subject, since most of such analyses have been conducted on individual vehicle units, not on networks as a whole. In addition, evaluation of road network reliability is mostly applied to general traffic but few studies have been conducted in relation to risk assessments and reliability analyses of risk-inherent traffic or hazardous material transportation.
Analysis on hazardous material transportation
239
CURRENT ANALYSIS OF HAZARDOUS-LOAD TRANSPORTATION
Current analysis of hazardous-load transportation Category of hazardous materials. Among all load materials categorised by the Road Traffic Census, "volatiles" (e.g. gasoline and benzene), "other petroleum and petroleum products" (hereafter petroleum products) and "chemicals" can be regarded as general hazardous materials. This section reviews the transportation of these three categories and outlines a spatial analysis of transporting "petroleum products" in particular, which constitutes a considerably high traffic volume. Volume of hazardous-load traffic. The traffic volume relating to "petroleum products", "volatiles" and "chemicals" in Hokkaido amount to approximately 71 thousand TE/day, 17 thousand TE/day, and 7 thousand TE/day respectively. The dead-weights are approximately 143 thousand t/day for petroleum products, 29 thousand t/day for volatiles and 6 thousand t/day for chemicals.
Figure 2 Volume of hazardous-load transportation in Hokkaido (Left: Trip generated and attracted volume, Right: Load volume) Source: 1999 Road Traffic Census
As for the proportion of inter-municipal and intra-municipal movements of hazardous-load transportation, inter-municipal movements represent only 20-30% of the total number of vehicles. However, dieir load volumes amount to as much as 40-70% of the total volume.
240 Logistics systems for sustainable cities
Figure 3 Relative proportions of inter- and intra-municipal movements of overall hazardous-load transportation (Number of vehicles, Load volume) Source: 1999 Road Traffic Census
Although the proportion of hazardous-load vehicles using expressways accounts for a significantly low value of just 1 to 3% of all hazardous-load vehicles, the proportion of the total load volume is much higher. In particular, the proportion of volatiles transported on expressways is nearly 20% of the total value.
Figure 4 Proportion of hazardous-load vehicles using expressways (Number of vehicles, Load volume) Source: 1999 Road Traffic Census
Similarly, the proportion of hazardous-load vehicles travelling to and from Sapporo accounts for 10-20% of the total vehicle number, although the volume of load is somewhat higher particularly for chemicals, of which nearly 40% is transported to and from Sapporo.
Analysis on hazardous material transportation
241
Figure 5 Proportion of hazardous-load vehicles travelling to and from Sapporo (Number of vehicles, Load volume Source: 1999 Road Traffic Census)
The volume of traffic transporting petroleum products, which is observed in terms of OD pairs, shows a strong connection between port cities and large cities. In addition, broader flows such as those between Ishikari and Wakkanai as well as Hakodate and Asahikawa can also be observed.
Figure 6 Volume of traffic transporting petroleum products Source: 1999 Road Traffic Census When we focus on the load volumes of hazardous-load shipments, movements around port cities rather than large cities become noticeable and the flow along the Hidaka area also becomes apparent.
242 Logistics systems for sustainable cities
Figure 7 Volume of transported loads for petroleum products Source: 1999 Road Traffic Census
Accident occurrence risk for hazardous-load vehicles Calculation and analysis requirements. A certain proportion of vehicles transporting hazardous substances such as inflammable and poisonous materials are included in overall road traffic. No matter whether they become primary or secondary accident vehicles, they are capable of causing extremely large detrimental effects to surrounding natural and social environments once involved in a traffic accident. From the vehicles listed in overall traffic accident and road traffic census data, road tankers (a special purpose freight vehicle equipped with a tank) were selected as hazardous-load vehicles, and a cross tabulation based analysis was conducted on accidents involving such vehicles on national highways in Hokkaido between 1990 and 1998. Analysis results. Analytical results of the characteristics of traffic accidents which involve hazardous-load vehicles are as follows. 1) Traffic accidents which have taken place on national highways within Hokkaido Road conditions More traffic accidents occur on rural coastal roads connecting major cities as well as mountain passes. Non-road tanker related accidents mainly take place on dry road surfaces (nearly 60% on dry road surfaces in contrast to only 20% on icy road surfaces), whereas accidents actually involving road tankers occur at a high rate of 40% on frozen road surfaces. Weather conditions Areas susceptible to accidents tend to be roads which are often subject to temporary closure in winter due to reduced visibility during snowstorms. The number of accidents which occur during periods of low visibility due to snowstorms is relatively high.
Analysis on hazardous material transportation
243
Figure 8 Number of accidents involving road tankers (over a nine year period)
Figure 9 Relative rates of types of winter accidents
Figure 10 Relative rates of accident occurrence by road surface condition
244 Logistics systems for sustainable cities
Figure 11 Number of accidents by road sections (in nine years)
Figure 12 Accident rate by road sections (cases/100 million vehicle km)
Analysis on hazardous material transportation
245
Accident types Traffic accidents involving hazardous-load vehicles are mostly caused by other vehicles. In terms of accident types, rear-end collisions at intersections account for 60% of the total number of accidents where a hazardous-load vehicle is the primary accident vehicle, whereas head-on collisions on curved road sections with an on-coming vehicle which has deviated from its own lane is the major type of accident implicating hazardous-load vehicles as secondary accident vehicles.
Figure 13 Accident types
Figure 14 Type of accidents occurring at road curves
2) Accidents which occur in central Hokkaido Road conditions A higher frequency of accident occurrence characteristically takes place on curved road sections and comparatively straight road sections with small longitudinal gradient. Many accidents are registered in Sapporo, the capital city of Hokkaido, as well as in Tomakomai and Muroran, Hokkaido's two major industrial cities. The number of accidents is particularly high around Tomakomai, where petroleum storage facilities are located, from which volatiles are transported. Table 3 Characteristics of hazardous-load vehicles and conditions which elevate their risk of accident occurrence Factor Road section Road section and condition
Road environment
Road location Road surface condition Season Weather Time
Accident caused a hazardous-load vehicle Rear-end collision at an intersection
by
Accident implicating a hazardous-load vehicle Head-on collision at a curved section
Non-urban district
Non-urban district
Frozen road surface
Frozen road surface
Winter Snowfall (reduced visibility) Daytime
Winter Clear sky Daytime
Conclusion and future tasks. Our findings after completing the process of identifying the attributes of traffic accidents which involve hazardous-load vehicles are that weather and road conditions are closely related to accidents in which hazardous-load vehicles are the primary accident vehicle, and that accidents involving hazardous-load vehicles as secondary accident vehicles are more often than not caused by other vehicles on certain types of road structure. For future risk evaluation of hazardous-load transportation, the establishment and formulation of parameters is required which take into account road structure, roadside environment, traffic volume and annual seasons. Based on these parameters a new method capable of identifying and evaluating accident occurrence risks for a selected route then needs to be studied and developed.
246
Logistics systems for sustainable cities
STUDY FOR RISK EVALUATION METHOD FOR ROAD NETWORKS Background and objectives Volatiles such as gasoline are hazardous materials in themselves, and therefore pose a risk of adversely affecting surrounding areas when vehicles transporting them are involved in accidents on the way to their intended destination, hi an effort to reduce these risks posed by hazardous-load vehicles, we focused on two strategies: 1) choosing less populated areas as optimum routes for the transportation of hazardous materials; and 2) completing transportation operations in a short time by choosing the shortest route.
Risk evaluation model for hazardous-load transportation Prerequisites for a risk evaluation model. If transported hazardous materials are unintentionally released due to an unexpected accident or the container is damaged on the way to their destination, some materials are likely to inflict a certain degree of damage to their immediate roadside environment. The assumption in the risk evaluation model of this study is that hazardous materials being transported are released along all sections of a selected route and some sorts of hazardous loads inflict a certain degree of damage on all roadside environments. Formulation of a risk evaluation model. It is generally believed that the adverse effects of hazardous materials if released are concentrated within a certain range of a area, but also that their effects become gradually reduced over a distance until no effect is any longer observed. Yet, to prepare for an unexpected worst-case scenario or to enhance awareness of risk management among roadside residents, the possibility of potential damage even in areas distant from a possible spill site should not be neglected for the sake of assured safety. By taking the above assumption into account, the following exponential function was adopted to estimate damage probability. This exponential function is capable of determining a low probability, but not zero, even for a remote location. (1) where, P(d)- Probability of damage occurrence at a distance (d) from transported hazardous materials d: distance from transported hazardous materials „ : damage damping parameter subject to the hazardous material type.
Analysis on hazardous material transportation
247
Figure 15 Change in probability reduction with different „ values (exponential function) Definition of a damage risk. A method to quantitatively calculate the possible degree of damage inflicted by a given hazardous-load shipment on the roadside environment in a selected route was studied. To find the route in which the roadside environments incur the minimum possible damage, the total number of estimated exposure victims in all areas are summed up for all possible incidents along the route and the one with the smallest total is determined as the route of minimum damage. This can be expressed as (damage risk) = (the number of potential victims)x(travel distance), and equation (2) formulates the damage risk for which "person-distance" is given as a unit.
(2) where, R(Zd)~- Damage risk over an OD distance (Zd). Nm: Population of a city (m). P(dm): Probability of damage occurrence in a city (m), which is determined by the distance (dm) between transported hazardous materials and the city. dm: Distance between transported hazardous materials and a city (m). Zo'- Origin of transported hazardous materials. ZQ- Destination of transported hazardous materials.
Application of a Risk Evaluation Analysis to an Actual Network Selection of a network. To build an evaluation framework, only the national highway network was extracted from all the road networks, to which locations of intersections and municipal government buildings were added as nodes (their co-ordinates were determined using GIS software). Then, the population for each municipality was incorporated into the system. Route with the minimum damage cost. Definition of the shortest-distance route If the travel speed and the degree of congestion between origin and destination are constant, the shortest-distance route should be selected to minimise the transportation time and cost. Such a (shortest-distance) route can be obtained by Dijkstra's algorithm.
248 Logistics systems for sustainable cities Definition of the least-populated route The least-populated route is the one in which the roadside population is the lowest among all possible routes. The link population between nodes can be calculated in equation (3), to which Dijkstra's algorithm is applied to obtain the least-populated route.
(3) ltf. Link population between nodes (i) and (j) Nt, Nj: Population at nodes (i) and (j), respectively
Risk in a unit distance. To determine a damage risk for each link using an equal standard, a damage risk (person/km) in a unit of distance was calculated in equation (4).
(4)
where, Rtj : Damage risk per unit distance between nodes (i) and (j). Zt '• route distance from an origin 0 to a node i Z
'• route distance from an origin 0 to a nodey
Comparison of the shortest-distance route and the least-populated route. An assumption that hazardous-load transportation should select the least-populated route rather than the shortest-distance route to minimise risk sounds reasonable. However, because the damage risk is the product of the total travel distance and the expected number of exposure victims, the damage risk of the least- populated route may exceed the shortest-distance route if the total travel distance increases beyond a certain figure. To settle the above question, the Nanyo area in Shikoku was used as a model to which these route selection methods were applied. Figure 16 shows the correlation between the total travel distance and the damage risk. According to the analytical results, no large distribution differences were recognised between the shortest-distance route and the least populated route within a total travel distance of 60km. Beyond a distance of 60km, however, the damage risk in the shortest-distance route became greater than that of the least-populated route. The least-populated routes with a damage risk of over 350 thousand person-km are plotted where the total travel
Analysis on hazardous material transportation
249
distance exceeds 120km, which suggests that the damage risk in the least-populated route becomes greater when the travel distance is longer than 120km. Findings. The following are the outcomes and findings of the analysis conducted on the Nanyo area model to obtain a minimised cost (risk) route. • For total travel distances within 60km, the optimum route for hazardous-load transportation differs depending on OD pairs. • As Figure 17 indicates, the shortest-distance route is also identified as the least-populated route in 48% of all the OD pairs. • Among all the OD pairs, 24% of them are optimised by the shortest-distance route and 28% are optimised by the least-populated route. • From the above outcomes, it can be seen that the route of minimum risk depends upon the OD pairs.
CONCLUSION The objective of this study was to establish a risk evaluation method for the transportation of hazardous-loads. Results obtained in the earlier sections are summarised as follows. First, a general formula was proposed to model the risk incurred during transportation of hazardous materials. The effects induced by hazardous materials depend upon the distance, and the number of exposure victims is given as an expected value. The risk evaluation model for transporting hazardous-loads can be expressed as the product of the expected number of exposure victims and the total travel distance. Second, the risk evaluation model was applied to the road network in the Nanyo area of Shikoku. As a method to minimise risk, two possible optimum routes, the shortest-distance route and the least-populated route, were proposed and analysed with the risk evaluation model. Optimum minimum-risk routes for hazardous-load transportation in the Nanyo network differ from OD pair to OD pair if the total travel distance is less than 60km. When the total travel distance is between 60km and 120km, the selection of the least-populated route produces a lower risk. On the other hand, the level of risk posed by the shortest-distance route becomes lower (than that of the least-populated route) when the total travel distance becomes 120km or longer. Therefore, the optimum route to minimise risk depends upon the nature of the OD pairs. Although several problems have been identified in determining the link population and the number of nodes, these are issues we shall try to address in the future.
REFERENCES Frank, W.C, J.C. Thill and R. Batta (2000). Spatial decision support system for hazardous material truck routing. Transportation Research, Part C8, pp.337-359. Infrastructure Planning Committee (1998). Kotsu Network no Kinko-Bunseki (Equilibrium analysis of traffic network). Saishin-no Riron-to-Kaiho. Japan Society of Civil Engineers.
This page is intentionally left blank
18
ANALYSING THE POTENTIAL IMPACTS OF SUSTAINABLE DISTRIBUTION MEASURES IN UK URBAN AREAS
Julian Allen, Transport Studies Group, University of Westminster, London, UK Michael Browne, Transport Studies Group, University of Westminster, London, UK Graham Tanner, Transport Studies Group, University of Westminster, London, UK Stephen Anderson, Transport Studies Group, University of Westminster, London, UK Georgina Christodoulou, Transport Studies Group, University of Westminster, London, UK Peter Jones, Transport Studies Group, University of Westminster, London, UK
ABSTRACT This paper presents a project that has investigated the current freight transport operations of seven different companies in three urban areas in the UK. The potential operational, financial and environmental effects of four policy measures on these operations were investigated by obtaining each company's expected response to these policy measures and applying these changes in their behaviour to their current operational vehicle data. The results suggest that the policy measures will vary in their operational and financial impact on the distribution companies and also in terms of the change in vehicle pollutant levels. Low Emission Zones are expected to have the least impact of the measures studied on the organisation and operation of distribution activities, but would have a potentially significant impact on pollutant levels. Time and weight restriction scenarios are predicted to have the greatest impact on the cost of distribution operations and may increase the environmental impacts of distribution activities, depending on how companies behave in response to such measures.
INTRODUCTION The University of Westminster has recently completed a research project that has examined the current urban freight activities of distribution companies in different parts of the UK
252
Logistics systems for sustainable cities
distribution market, in terms of their operational, financial and environmental impacts. The project has also investigated the likely effect of transport policy measures that could be introduced in the medium term on these operations. The project was carried out in collaboration with Birmingham City Council, Hampshire County Council, Norfolk County Council and several distribution and logistics companies. It was concerned with the sustainability of current urban freight operations and the ways in which this would potentially be altered by company responses to policy measures. The paper firstly describes the context in which the project took place. It then describes the project methodology and presents results of the analysis of current urban freight operations. Finally, the policy measures investigated and the results of the companies' likely reactions to these policy measures are discussed.
BACKGROUND TO THE PROJECT While the recent growth of research into urban distribution is encouraging (see for example Ambrosini et.al., 2001; Kohler, 1999; Meimbresse and Sonntag, 2000; Thompson and Taniguchi, 2000), little of this work has been concerned with examining the likely impact of policy measures on distribution operations. Few previous studies have attempted to understand the relationship between: (i) policy measures, (ii) likely company action in response to the measure (in terms of distribution activity), (iii) the effect on operating costs, and (iv) the change in environmental impact. The intention of our project was to: • • •
•
Illustrate patterns of current urban distribution operations of different companies and the variation between them; Show the extent to which distribution operations vary for the same company in three different urban areas studied; Quantify the likely direction and scale of change in distribution operations, vehicle operating costs and environmental impacts for different patterns of distribution if new policy measures were introduced in urban areas; Indicate whether the policy measures are likely to result in the same or different outcomes in the three urban areas studied.
The main aim of the project was to investigate the ways in which policy measures are likely to result in changes in goods vehicle activity for different types of urban distribution operation. Policy measures tested include Low Emissions Zones, congestion charging, weight and access time restrictions. Changes that companies could make to their operations in order to improve efficiency while at the same time reduce environmental impacts were also considered in the project but are not reported on in this paper. This project has built on our earlier work, further developing our work into the sustainability of urban distribution (Allen et al., 2000).
RESEARCH APPROACH The research approach comprised the following four main components: i. Devising a suitable assessment approach to examine the relationships between policy measures, companies' distribution operations, and financial and environmental impacts of these operations;
Analysing sustainable distribution measures
253
ii.
Development of a database model to reflect the relationships noted in (i) above. This model was used to calculate operational, financial and environmental indicators of companies' current distribution operations. The database model has also been used to reflect the change in these indicators caused by likely alterations to distribution operations resulting from the introduction of various policy measures; iii. Working closely with distribution companies that carry out urban collections and deliveries in order to firstly, understand and document their existing goods flow and vehicle activity patterns in the three urban areas studied. And secondly, to ascertain how these companies and their customers would expect these patterns of operation to change as a result of the introduction of specific policy measures; iv. Evaluation of the similarities and differences between the distribution operations and the three urban areas studied in the project in terms of: (a) the current pattern of goods collections and deliveries of the distribution companies, (b) the likely change in the goods collection and delivery operations in response to potential policy measures, and (c) the extent of change in the operational, financial and environmental indicators as a result of the new pattern of goods collection and delivery operations. Urban distribution operations were studied in Birmingham, Basingstoke and Norwich. These locations were chosen due to their differences (in terms of scale, age, urban form, and geographical location) and also because of the enthusiasm of these local authorities to participate in a project that would help them to better understand current patterns of operation, and the relationships between policy measures, distribution operations, and environmental impacts. We studied seven distribution operations in detail. This allowed us to reflect several major patterns of urban distribution operation in the project including parcels delivery, general haulage, the distribution of drinks, and dedicated contract distribution for retail stores. Figure 1 shows the different activities in the project and how they relate to each other.
254 Logistics systems for sustainable cities
Figure 1 The project activities The project involved collecting a significant amount of data from the distribution companies. This included: (i) a detailed three-day survey of vehicle rounds in the three urban areas, and (ii) a more general survey of the total distribution activity taking place at the depots from which these vehicles operate. Detailed information about 120 vehicle rounds carried out by the companies was collected. In total, 2286 collections and deliveries were made on these rounds. The database developed and used in the project had to be capable of handling all the distribution data collected. It also had to be designed to allow manipulation of this data to reflect operational changes resulting from policy measures in accordance with the views expressed by companies. A set of indicators was selected to reflect the sustainability of these current vehicle rounds. These included: (i) operational indicators (including 'time taken', 'speed', 'distance travelled', 'vehicle fill', and 'proportion of on- and off-street deliveries'), (ii) financial indicators (based on the cost of making deliveries and collections to the distribution company) and (iii) environmental indicators (including CO, CO2, NOx and PM10 emissions). Although 'time taken', 'speed' and 'distance travelled' are listed above as operational indicators, they also obviously influence the environmental impact of distribution activities.
Analysing sustainable distribution measures
255
POLICY MEASURES INVESTIGATED The task of determining which policy measures to test in the project was carried out in conjunction with the local authority partners from Birmingham City Council, Hampshire County Council, and Norfolk County Council. Three basic principles were used to guide the selection process: • • •
To assist in making comparisons it was agreed that the same policy measures would be tested in each urban area; Policy measures selected should be area-wide approaches to goods vehicle management in an urban area, rather than localised measures in a particular street; Measures selected were thought to be the types of approach that policy makers may take with respect to UK urban areas in the next five years (i.e. policy measures that may be implemented in the medium term).
Four policy measures were selected for detailed analysis: •
Low Emission Zones - the aim of a Low Emission Zone (LEZ) is to improve air quality by excluding older, high-polluting goods vehicles from certain urban areas and encouraging the faster take up of more modern, cleaner vehicles. Such zones do not currently exist in UK towns and cities;
•
Congestion charging - this refers to a scheme in which vehicle drivers (or the companies responsible for the vehicles) have to pay a charge in order to enter a particular geographical area at a particular time. The aim of such a scheme is to reduce road traffic levels in the urban area and also to reduce traffic pollutant emissions. Such a scheme may also generate a profit which can be used to provide improved public transport services. Congestion charging was introduced in London in February 2003 but does not exist in any other UK urban area;
•
Vehicle weight restrictions - in this policy measure only vehicles below a specified gross vehicle weight would be allowed to enter a specific geographical area within the urban area to make collections and deliveries during a large period of the working day (10:00 to 16:00). The aim of such a measure would be to reduce the number of large goods vehicles entering the chosen area when pedestrians and other road users are present and thereby overcoming the impacts that it is commonly perceived that these vehicles cause, such as pollution, intimidation, safety concerns, vibrations and noise;
•
Vehicle access time restrictions - in this policy measure no goods vehicles would be permitted to enter a specified geographical area within the inner urban area to make collections and deliveries during a large period of the working day. The aim of such a measure would be to prevent goods vehicles of any weight entering the chosen area when pedestrians and other road users are present. This could help to reduce the impacts that it is commonly perceived that goods vehicles cause, such as pollution, intimidation, safety concerns, vibrations and noise.
Four scenarios were tested for each policy measure. The next stage involved determining how the distribution companies expected these policy measures to affect their operations in the three urban areas (Birmingham, Basingstoke and Norwich). To achieve this a meeting was held with the distribution companies at which the company representatives were presented with an
256
Logistics systems for sustainable cities
explanation of the policy measures and the different scenarios for that measure, and then asked to comment on and discuss the likely company response to the scenarios for each policy measure. Detailed discussions enabled the representatives to describe likely changes that the company would make in order to meet the specific scenarios put to them. Companies were also given the opportunity to explain initiatives they could implement that would result in operational, financial and environmental benefits. The companies' responses to the policy measures were modelled by applying queries to current distribution data in the database using relationships derived from company interviews. Each scenario was treated in this way and by doing this revised distribution data was calculated for each vehicle round studied. The difference between the current and revised indicators was then calculated for each scenario. Figure 2 shows the relationship between data items in the database used for producing these indicators.
RESULTS Current operations The operations studied ranged from multi-drop work with 100 deliveries per round made with light goods vehicles from a depot located within the urban area, to full-load single drop work on large articulated vehicles made from a single national distribution centre. The project has
Analysing sustainable distribution measures
257
quantified key differences in the current urban distribution operations of the companies taking part in the project. A range of local, regional and national vehicle round patterns were observed among the seven distribution companies studied in the project. These are listed below. •
Urban distribution operation (i.e. vehicle rounds operated from depot located in urban area): Collections/deliveries wholly within city centre; Collections/deliveries wholly within rest of urban area (i.e. not in city centre); Collections/deliveries in city centre and rest of urban area; Collections/deliveries in urban area and outside urban area.
•
Regional distribution operation (i.e. vehicle rounds operated from depot located in same region as urban area): Collections/deliveries wholly within one urban area; Collections/deliveries in more than one urban area.
•
National distribution operation (i.e. vehicle rounds operated from national depot to the urban area): Full load delivery for one destination in urban area; Collections/deliveries wholly within one urban area; Collections/deliveries in more than one urban area.
The work has demonstrated that there are important differences between current urban distribution operations that need to be understood when considering urban distribution policy measures and their likely effects. Potential impact of the four policy measures The four policy measures analysed were found to have differing effects on distribution activity in terms of: (i) impact on how the distribution operation would be carried out and its cost, and (ii) on the environmental impact of the distribution operation. In addition, results varied by type of distribution operation and by urban area. Unfortunately, too many scenarios were modelled to include detailed results in this paper, however the findings for each policy measure and differences between companies and urban areas are summarised in the following sections. The full results are available in the final project report (Allen et al., 2003). An example of the database outputs is shown in Tables 1 and 2, which contain the results by urban area and by company for the LEZ policy measure applied to the seven companies' vehicles operating in the inner urban area.
258 Logistics systems for sustainable cities
Analysing sustainable distribution measures
259
Operational, financial and environmental impacts Low Emission Zones The results indicate that Low Emission Zones will have the least impact of the four policy measures on the organisation and operation of distribution activities, but would have a potentially significant impact on pollutant levels (for example a LEZ based on Euro III engine standards could lead to reductions in paniculate matter of up to 50% for the companies studied). However, three of the seven companies would be expected to experience vehicle operating cost increases of up to 5% due to the need to acquire compliant vehicles. Operating costs and environmental impacts will depend on two main factors: (i) the company's vehicle replacement cycle and (ii) the geographical profile of a companies delivery and collection work compared with the geographical coverage of a LEZ scheme. Four LEZ scenarios were investigated (based on different geographical areas and engine requirements) and their effects assessed for different types of distribution operation. Congestion Charging Our work has demonstrated that the effect of congestion charging will differ between companies, depending on: (i) the level of the charge, (ii) the geographical area in which the scheme is implemented, and (iii) whether or not the scheme results in speed improvements. The results suggest that improvements in the average speed of goods vehicles (as a result of reductions in traffic levels) can reduce, and in the case of some companies outweigh, the congestion charge (depending on the level of the charge). The findings indicate that a 15% reduction in driving time in the congestion charging area, would more than offset a daily congestion charge of £5 per vehicle for some companies. However a daily charge of £15 would lead to increased operating costs for all companies. The work has highlighted the importance of generating time savings to ensure that congestion charging does not have a detrimental financial effect on distribution operations, in helping to increase acceptability among companies, and for pollution reduction to be achieved. Weight Restrictions The companies studied would be affected very differently by weight restriction policy measures. In the scenarios we have examined, those companies operating light goods vehicles (i.e. up to 3.5 tonnes gross weight) would be completely unaffected, while those companies operating heavy goods vehicles with a gross weight of 12 tonnes or more would have to make significant changes to their distribution patterns in order to comply (i.e. operating a greater number of vehicle rounds using lighter vehicles) These changes would result in increases in total vehicle operating costs of as much as 30% for some companies depending on the weight restriction. The environmental impact of the vehicle rounds performed by those companies worst affected by the weight restriction scenarios would increase significantly as a result of increases in total distance travelled (calculated to double for one company if a 7.5 tonne gross vehicle weight limit was introduced), which would lead to increases in total fuel consumption and pollutant emissions. The increase in the total time taken to complete the same quantity of collection and delivery work (which is expected to rise by as much as 50% in the case of one company) would also lead to negative impacts. Time Restrictions Time restrictions could lead to distribution activities being compressed into a shorter period at the start or end of the working day. If this were to happen the results suggest that, like weight restrictions, there would be negative impacts on the distribution operations of companies affected in terms of increases in vehicle rounds, total distance travelled and could lead to more
260 Logistics systems for sustainable cities queuing at receivers' premises. The environmental impact of vehicle activity would also increase if companies responded to time restrictions in this manner. However, if time restrictions resulted in more distribution companies operating at night then the results indicate that this could be beneficial from both a commercial and environmental perspective. The commercial benefits would depend on the trade-off between improved driving speeds and higher drivers' wages. The results indicate that improved driving speeds due to night working could result in vehicle operating cost reductions of between 1-4% for the companies studied, as well as reductions in pollutant emissions. Though it must be recognised that there may be noise implications for local residents. However, if drivers' wages were to rise by 20% for night working this would outweigh the value of improved driving speeds and would lead to operating cost increases of 1-4% for the companies studied. For night delivery and collection to become more commonplace it would be necessary for senders and receivers of goods to accept night work. They will potentially experience higher reception/despatch costs and may have concerns about the safety of their premises if staff were not present. Therefore negotiations between supply chain partners would be necessary to make night collections and deliveries in urban areas possible for more distribution companies.
Impact of policy measures on different companies / types of operation The results indicate that the four policy measures (and the scenarios modelled for each) will not result in uniform effects for all urban freight transport operations. Differences between companies for each policy measures are presented in the following sections. Low Emission Zones The LEZ policy measures tested in this project are not expected to affect the vehicle operating costs of companies in particular sectors of the distribution market more than others. The impact on vehicle operating costs will depend on two main factors: (i) the company's vehicle replacement cycle, and (ii) the geographical profile of a company's collection and delivery work compared with the geographical coverage of a LEZ scheme. However, if LEZ schemes were only implemented in a small number of urban areas in the UK, large companies with a national fleet may be able to redeploy their newer vehicles to these urban areas and use the older vehicles in areas without a LEZ. Locally-based companies with small fleets would be unable to redeploy their vehicles in this way. The latter may therefore be disproportionately affected by a LEZ scheme in terms of the fleet changes they will have to make in order to comply. Congestion Charging The results suggest that some of the companies in the project will be more adversely affected than others by particular congestion charging scenarios. Very few of the vehicle rounds we studied avoid the congestion charge due to completing their work in the congestion charging area before, or starting their work after, the charge comes into force. However, like the LEZ policy measures, differences in the impact on companies are due to the proportion of a company's vehicle rounds that take place in the geographical area in which the scheme is implemented.
Analysing sustainable distribution measures
261
A congestion charging scheme that was implemented in the central/inner area may disproportionately affect distribution companies delivering to the city centre such as parcels carriers, and companies delivering to high street shops and to pubs, bars and restaurants. If the congestion charging scheme covers the entire urban area then many companies would be expected to have a high proportion of affected vehicle rounds. Weight restrictions The results indicate that the weight restriction policy measure would be expected to affect the companies studied in the project very differently. Some would be totally unaffected while others would have to make significant changes to their operation, and would experience sizeable cost increases. The primary factor in determining how severely distribution companies would be affected by such weight restriction measures is obviously the weight of vehicles currently operated by companies. Other important factors are the geography of vehicle rounds compared with the area covered by the weight restriction, and the times at which vehicle rounds take place compared with the times at which restrictions are in force. Time Restrictions Three of the companies studied would have a greater proportion of vehicle rounds affected by the time restriction scenarios modelled in the inner area than the other companies. This is due to two facts: (i) that they have a high proportion of rounds that enter the inner area, and (ii) that many of their rounds take place during the restricted times of 10:00 to 16:00. Differences in current and predicted operations in the three urban areas The project has demonstrated that the size and form of the urban area has an important bearing on the distribution operations that serve that area. For example, average speeds in Basingstoke are higher than in Birmingham and Norwich, and the proportion of off-street and shopping centre deliveries are higher in Basingstoke (which was designed in the 1950s and 1960s to segregate and improve much of the distribution work for the city centre). Multi-drop vehicle rounds serving Basingstoke also cover greater distances than in Norwich and Birmingham as, given the lower population density, it is necessary to travel further to carry out the same amount of collection and delivery work. Discussion with the companies together with our own analysis of each urban area has shown that existing distribution problems (both in terms of problems caused and experienced by goods vehicle) are far more acute in Birmingham and Norwich than in Basingstoke. The work has also shown that in some cases the policy measures studied would be likely to either: (i) affect a different proportion of vehicle rounds in the three areas, or (ii) result in different company reactions in the three areas, thereby resulting in different environmental outcomes.
262
Logistics systems for sustainable cities
CONCLUSIONS Even with the relatively small number of companies participating in the project and the amount of vehicle round data captured, it has been possible to obtain much insight into the likely company reactions to and effects of different policy measures. This work will add to the current discussion about policy making for urban distribution. Within the current research project, extensive use has been made of the distribution database compiled from vehicle round data but there is a scope to extend this by collecting additional data about vehicle rounds in other urban areas.
REFERENCES Allen, J., G. Tanner, M. Browne, S. Anderson, G. Christodoulou and P. Jones, (2003). Modelling policy measures and company initiatives for sustainable urban distribution — Final Technical Report, project carried out as part of the EPSRC/DfT Future Integrated Transport Programme, University of Westminster, http://www. wmin.ac.uk/transport/projec ts/sus_u-d.htm Allen, J., S. Anderson and M. Browne and P. Jones, (2000). A framework for considering policies to encourage sustainable urban freight traffic and goods/service flows Summary Report, project carried out as part of the EPSRC Sustainable Cities Programme, University of Westminster, http://www.wmin.ac.uk/transport/download/urbandistsumm.pdf Ambrosini, C, J. Routhier and D. Patier-Marque, (2001). Objectives, methods and results of surveys carried out in the field of urban freight transport: an international comparison, paper presented at 9th World Conference on Transport Research (WCTR) in Seoul: Korea, 22-27 July 2001. Kohler, U. (1999). City Logistics in Kassel, Proceedings 1st International Conference on City Logistics, 12-14 July 1999, in Cairns: Australia, (E. Taniguchi and R.G. Thompson, eds), Institute for City Logistics, 261-271. Meimbresse, B. and H. Sonntag, (2000). Modelling Urban Commercial Traffic with the Model Wiwer, paper presented at Jacques Carder Conference, Montreal: Canada, 4-6 October 2000. Thompson, R.G. and E, Taniguchi, (1999). Routing of Commercial Vehicles Using Stochastic Programming, Proceedings 1st International Conference on City Logistics, 12-14 July 1999 in Cairns: Australia, (E. Taniguchi and R.G. Thompson, eds), Institute for City Logistics, 73-83. The project was led by Professor Michael Browne, Head of the Freight Transport and Logistics Unit in the Transport Studies Group (TSG) at the University of Westminster. The authors would like to thank the Department for Transport and EPSRC for the funding that the project received as part of the Future Integrated Transport programme.
19
FUTURE CITY LOGISTICS IN JAPAN FROM THE SHIPPERS' AND CARRIERS' VIEW - PROSPECTS AND RECENT MEASURES TO DEVELOP THEM
Katsuhiko Hayashi, University of Marketing and Distribution Sciences, Kobe, Japan Yuji Yano, Ryutsu Keizai University, Ryugasaki, Japan
ABSTRACT This paper discusses the views of shippers and carriers regarding the future prospects and likely direction of city logistics in Japan, based on a survey conducted by mail. It also offers some concrete examples of how companies are making their logistics systems more efficient and environmentally sensitive, based on policy papers prepared by the companies themselves. The paper examines their positions and attitudes toward hypotheses on issues such as the environment, government policy, physical distribution management, and information technology, and discusses some the most important issues. The trade-off between environmentfriendly logistics and quality of logistics services - one of the issues that respondents viewed as incompatible — will be a key challenge for city logistics. The paper offers some examples of how this trade-off can be addressed, based on the policy papers on the environment published by many companies. Some companies are now trying to make their distribution systems more efficient and environmentally sensitive, while also introducing measures such as a review of distribution systems, precise transport management, the use of back hauling, and collective delivery systems.
INTRODUCTION When considering city logistics issues, it is important to take particular note of the implications for shippers and freight carriers, and how these will affect their efforts in coping with future
264
Logistics systems for sustainable cities
changes. Shipping companies and freight carriers in Japan are currently facing severe pressure to reduce the environmental impact of their logistics activities as well as to make them more efficient. In most of Japan's urban centres, there has been a chronic failure to meet environmental standards regarding to the nitrogen dioxide (NO2) and suspended particulate materials (SPM). Exhaust gases from diesel trucks have been identified as one of the major pollutants causing this problem. Local governments are now introducing more stringent rules on truck emissions. Logistics activities need to be made more environment-friendly, despite the expenses that this will entail. Meanwhile, as shipping organisations try to decrease their inventories, they have a tendency to make deliveries in smaller lots, more frequently. This tendency as well as fierce competition between carriers, leads to increasing truck traffic. Stagnation in the Japanese economy over the past several years has increased the pressure to cut logistics costs. The gradual deregulation has made the logistics market more competitive. Recent developments in logistics management may make the situation even more complex. As more shippers apply the concept of supply chain management (SCM) to their logistics operations, carriers and forwarders have begun to develop more sophisticated logistics services, with some adopting third party logistics (3PL). These developments could lead to more frequent deliveries and even more congestion, or they might provide a clue to resolving the trade-off between logistics and the environment. Useful information can be gleaned from a survey of shippers and carriers, to determine their perspectives on future prospects for city logistics and the rapidly changing business environment. This paper discusses the future prospects for city logistics in Japan, based on the opinions offered by shippers and carriers, as well as their recent activities. Our analyses is based on a survey conducted by the Japan Institute of Logistics Systems (JILS), a national non-profit organization specializing in logistics representing shippers and carriers, in 2001, in which we took part, as study group members. Numerous hypotheses were constructed regarding future trends in logistics. Respondents were asked to indicate which of the hypotheses, in their opinion, would come true by around 2010. They were also asked to indicate whether their company would take measures to address the situation, if the hypothesis did come true, and to what extent the situation would affect their industry. We also have attempted to develop a concrete model of how companies balance the trade-offs between environmental problems and sophisticated logistics activities. For this purpose, we have analyzed the environmental policy papers published by many companies and discuss their future logistics plans.
RESPONDENTS Questionnaires were sent to 1,015 members of JILS in April 2001. Almost all of the major shippers and carriers were included in the list. Of these, 309 questionnaires were returned, for a response rate of 30 percent. The individuals surveyed were all managers or heads of their company's logistics department.
Future city logistics in Japan
265
The respondents were categorised into separate groups — shippers and carriers — as shown in Table 1. A large percentage of the shippers were product manufacturers, reflecting the membership of the JILS. Other than manufacturers, this group included wholesalers, retailers, construction firms and others. Apart from shippers and carriers, there were respondents from the financial services, real estate and consulting industries and the government sector. A breakdown of respondents by the annual revenue of their organisation is provided in Table 2. The average size of shippers was far larger than carriers. Average annual revenues for responding companies amounted to 411 billion yen. Compared with the average revenues for Japanese companies in general, the respondents were from relatively large companies. Table 1 Respondents by industry Industry Shippers (Manufacturers) Carriers Subtotal Others Total
Number Percent 46% 141 (33%) (102) 39% 120 84% 261 48 16% 100% 309
Table 2 Respondents by annual revenue Revenue (billion yen) Less than one 1 - 10 10-100 100-1,000 More than 1,000 Not answered Total
Shippers 3 7 32 71 24 4 141
Carriers 10 52 40 13 3 2 120
PROSPECTS FOR FUTURE CITY LOGISTICS
Environment Background. Environmental impacts have become the most pressing issue for city logistics systems in Japan. Many people who live along major urban road arteries suffer from chronic environment-related health problems. The majority of monitoring posts located in cities show that nitrogen oxide (NOx) and suspended paniculate materials (SPM) levels exceed the amount permissible under current environmental standards. Some groups of patients who suffer from ailments related to exhaust gases caused by the heavy traffic have sued the local governments and road administrators, hoping to force action to meet environmental standards. In cases filed in NishiYodogawa, Kawasaki, Amagasaki, south of Nagoya and Tokyo, the governments lost the lawsuits, and have promised to take emergency action to satisfy the air quality standards. Since diesel trucks are major sources of NOx and SPM, governments are tightening the regulations on diesel trucks. They are also strengthening emission standards for diesel cars, and taking steps to control diesel truck traffic. "The law concerning special measures for total emission reduction of nitrogen oxides and paniculate materials from automobiles in specified areas" (NOx and PM from automobiles law) in 2001 sets forth the fundamental policies and plans for reducing the total volume of auto traffic. However, the transport of goods in urban areas is heavily dependent on diesel trucks, at present, because they are more economical than trucks with gasoline engines. It is not easy to convince private companies to replace diesel trucks with cleaner trucks, given the recession. Global warming is also an issue that needs to be tackled by shippers and carriers. Greenhouse gases such as carbon dioxide (CO2) must be reduced under international cooperation treaties. Japan's government has promised to reduce the amount of CO2 gas emissions by six percent
266
Logistics systems for sustainable cities
from 1990 levels, by 2010. Emissions of CO2 produced by logistics activities account for about one third of all such emissions in Japan and are still increasing. More efficient means of transporting cargo are needed. Recycling is becoming an important issue for Japan, as well, in order to ensure healthy and comfortable living conditions for both present and future generations. This is a particularly crucial issue for city logistics to address, since big cities have limited capacity for waste disposal. Recycling requirements were implemented in 2001, for televisions, air conditioners, refrigerators and washing machines, and they are likely to be extended to cover other goods in the future. Manufacturers of these four items are obliged to take back the used items and recycle them. Hypotheses and their prospects. Based on the background conditions described above, we generated several hypotheses regarding the future conditions affecting logistics systems and the environment, as shown in Table 3. We asked survey respondents whether they believed these hypotheses would come true, and if so, how they would cope with the changes. According to the majority of the respondents, environmental issues pose a serious challenge. Table 3 Hypotheses on environment and results Hypotheses about Future (aroundyear 2010)
Evaluations of Realization (*1)
Respondents' Directions to Cope •with the Hypotheses (*2) Shippers Carriers * 1.8 1.7
Shippers Carriers Harmonizing logistics systems with the environment 1.4 1.5 will become a matter of course for corporate policy. Environmentally sensitive logistics will take 1.6 0.5 1.5 0.3 precedence over quality of logistics service. ** ISO14001 certification will become an important 1.6 0.6 0.5 criterion for corporate evaluation. 1.4 Consideration of logistics and impact on the 1.2 4 environment (easy to carry, compact, etc) will be 1.7 1.1 1.7 ubiquitous in the design process. Packing and wrapping materials will be reduced, 1.6 1.7 0.8 1.0 reused, and almost completely recycled. 6 Low-emission tracks will replace diesel trucks. 1.7 1.6 1.2 "1.2 1.2 1.1 0.3 7 Tracks will be replaced by multi-modal transport. 0.3 1.6 1.6 1.2 1.1 8 Reverse logistics will become a big market. Regulation on recycling will spread to almost all 1.6 1.7 1.1 1.1 products. 10 NOx regulations will spread to all cities in Japan. 1.5 1.3 1.4 1.5 Exhaust gas regulations on trucks will be tightened to 1.6 1.6 1.5 1.5 reduce NOx emissions Regulations on transport in general will be tightened 1.6 1.5 1.4 1.3 to reduce CO2 emissions. Carbon tax will be introduced as one of several 1.1 1.1 0.3 0.3 environment tax systems. Notes: (*1) The scale used is as follows. Agree =2. Agree somewhat =1. Neutral =0. Disagree somewhat = -1. Disagree = -2. (*2) The scale used is as follows. The respondent's company will cope with the hypothesis proactively =2. Cope with unwillingly =1. Neutral =0. Impossible to cope with = - 1 . No intention to cope with =-2. (*3) The number of * indicates the significance of difference between shippers and carriers by Wilcoxon Scores (rank sums). *** lzKO.001 ** IzKO.Ol * IzKO.l
Future city logistics in Japan
267
Both shippers and carriers believe that, in the future, it will be a matter of course in corporate policy that logistics systems be harmonized with the environment. The respondents, especially shippers, replied that they will try to cope with the trends positively (Hypothesis 1). However, they did not agree strongly that green logistics will take precedence over the level of logistics service, although they will try to cope with the trade-offs (H2). It is no surprise that they expect environmental regulations, such as limits on NOx (H10, H l l ) and CO 2 emissions (HI2), to be strengthened, because governments already have plans to strengthen these regulations to cope with air pollution. Given this outlook, respondents are trying to cope with the regulations. Respondents agree somewhat that low-emission trucks will replace diesel trucks, and plan to make such replacements themselves (H6). However, they doubt that multi-modal transport will replace trucks (H7). They also take the view that controversial proposals, such as a carbon tax (H13), will not have been introduced by 2010. As for recycling, they generally agree that packing and wrapping materials will be reduced, reused, and recycled almost completely (H4) and that regulations on recycling will spread to cover almost all products (H5). As recycling continues, they agree to some extent that reverse logistics will become a big market and many plan to tackle this new business (H8).
Government policy Background. The Cabinet has adopted structural reform policies to revitalize the economy. In the field of logistics, the Cabinet adopted a comprehensive logistics plan in 1997 and revised it in 2001, to make logistics more effective and efficient. Under the revised plan, the Government will take comprehensive measures to make Japan's logistics market not only cost-competitive on a global basis, but also sustainable until 2005. There are many specific measures relating to city logistics. As for its policy of making logistics more competitive, the Government will ease economic regulations to derive the maximum benefits from the vitality of private entities. Regarding the issue of high logistics costs, the Government adopted a policy to selectively construct the important elements of distribution infrastructure. Public works projects aimed at decentralisation tend to be scattered among local areas without consideration of priorities, due to political pressure. Construction of logistics infrastructure in urban areas is unable to keep up with demand. New policies may improve the infrastructure of city logistics systems. In Japan, controls on city logistics activities are inevitable, in order to maintain environmental standards. Many local governments are considering controls on the movement of diesel trucks. The Tokyo metropolitan government is considering a plan to introduce road pricing, to reduce auto traffic. Hypotheses and their prospects. The respondents do not agree as strongly with the hypotheses regarding government policy as they do with the ones regarding the environment. They are especially sceptical about the hypotheses related to infrastructure. They neither agree nor disagree that a scattering of public works will be reconsidered (HI) and highway tolls will be decreased (H2).
268 Logistics systems for sustainable cities Although many cities such as Tokyo and Yokohama already carried out experiments with transport demand management (TDM), the respondents do not foresee that TDM will become a common means of utilizing infrastructure more efficiently (H3). They do not think road pricing will be introduced in many cities (H10). However, they think it probable that direct controls on heavy trucks will be tightened (H9). Concerning economic regulations, the respondents do not expect price and entry regulations to be abolished by 2010 (H5 and H6). On the other hand, they foresee tighter safety regulations, such as penalties for overloading and speeding (H7). In Japan, location regulations on logistics facilities have not been particularly strict in urban areas. As a result, facilities such as trucking terminals and warehouses are scattered. This tends to intensify traffic congestion. Respondents have a rather strong expectation that regulations on the location of logistics facilities will be tightened in urban areas and loosened in other areas (H8). Table 4 Hypotheses on government policy and results
Physical distribution management Background. City logistics in Japan is characterised by high quality service such as just-in-time distribution and frequent, small-lot deliveries. These high level delivery services have been criticised because they tend to cause traffic jams and can become a public nuisance if they are not systematised. Japanese consignees tend to demand higher quality distribution services because the costs of distribution are included in the price of the goods. In other words, consignees do not have to pay extra even if they ask for higher quality service. Thus, frequent small-lot delivery services are quite common, as stores try to decrease inventory levels while not missing sales
Future city logistics in Japan
269
opportunities due to stock-outs. However, the costs of such frequent delivery services are becoming a burden to shippers and carriers. They are trying to streamline their logistics systems to cut the cost of deliveries, while at the same time maintaining the level of service. Thus, physical distribution management is crucial for Japanese companies hoping to survive in today's competitive business environment. One key concept is supply chain management (SCM), that involves the integration of logistics processes from the supplier of raw materials, through manufactures, wholesalers, and retailers to consumers. Under the SCM, overlapping activities such as cross transport and redundant inventories are eliminated. Another important concept is third party logistics (3PL). In pursuing SCM, companies try to concentrate on processes where they can utilise their own core competences, and outsource other processes. Logistics is often one of the processes that is outsourced. 3PL services providers make blanket contracts to handle all logistics processes on the shipper's behalf. Hypotheses and their prospects. For all hypotheses, there were more respondents who stated that they agree or somewhat agree than there were those who disagree or somewhat disagree. However, regarding the degree of agreement, there was a big difference between hypotheses. Table 5 Hypotheses on physical distribution management and results Respondents' Directions to Cope with the Hypotheses Hypotheses about Future (aroundyear 2010) (*2) Shippers Carriers Shippers Carriers Level of logistics services will be viewed as an important part ** * 1.4 of differentiation strategy and various services will be 1.8 1.9 1.7 introduced. Shippers will pay more for excellent logistics services even 0.3 0.3 0.8 1.6 though they only pay attention to cutting costs, now. Traditional trading customs such as fringe services at no extra 1.4 0.5 1.3 -0.1 charge will disappear. Physical distribution will be separated from transactions. *** 1.8 1.7 1.2 0.5 Many companies will adopt the concepts and techniques of 1.1 1.7 1.7 SCM, and SCM will become common. 0.9 Retailers will form alliances with manufacturers and the role 0.7 1.2 0.6 1.5 of wholesalers will decline. Companies will form strategic alliances with others to 1.5 1.8 1.7 1.4 improve logistics efficiency. The outsourcing of logistics functions will become popular, 1.2 1.3 1.5 1.8 * # 0.7 Third party logistics will become a big market. 1.3 0.9 1.5 Reorganization and construction of distribution centres will ** 1.6 1.6 1.9 continue as part of shippers' logistics strategies. 1.7 Among the functions of physical distribution management, * * 0.4 1.4 transport will be more important. 0.7 1.5 JIT deliveries will be ordinary services for vendors. 0.9 0.9 1.4 1.4 The functions of storage warehouses and distribution centres 1.4 1.4 1.8 1.8 will become shipping oriented from storage oriented. (Notes: Same as Table 3) Evaluations of Realisation (*1)
1 2
,
, _ o 9 iri
12 .,
270 Logistics systems for sustainable cities The hypothesis which respondents viewed as most likely to be realised was the reorganisation and construction of distribution centres (H10). Logistics services as a corporate strategy (HI), alliances to improve logistics efficiency (H7) and shipping oriented distribution centres (HI3) were also viewed as highly probable. Following these hypotheses, many respondents agreed that logistics outsourcing (H8) would occur. In contrast, fewer respondents agreed that pricing changes (H2) or changes to traditional business customs (H3) would occur. These results indicate that carriers are likely to face more severe conditions. Shippers are not expected to increase compensation, even for excellent logistics services, even though they recognise the importance of logistics services and demand better services. Generally, there were few differences between the trend of responses for shippers and that of carriers. However, concerning traditional customs (H3), carriers tend to be more pessimistic about the possibility that the problem will be corrected, while shippers answered that it is likely to be corrected. Respondents' intentions regarding how to cope with the hypotheses were greatly influenced by their expectations about the likelihood. The correlation coefficient between the two is very high (r = 0.81). Each company answered that they will try to cope with the changes positively. The options selected by respondents as a way to cope with changes positively were as follows: logistics service as a corporate strategy (HI); alliances to improve logistics efficiency (H8); reorganization and construction of distribution centres (H10) and shipping oriented distribution centres (H13). Although there were few differences between the responses of shippers and those of carriers, the latter tended to focus more on resolving pricing problems (H2) and traditional customs (H3), as ways to cope with changes positively.
Information technology Background. Technical innovation can help to resolve the trade-offs among logistics systems. It is also expected to help ease the conflict between efforts to provide high level logistics services and environmental problems. Information and communications technology (IT), above all others, is expected to play a key role in future logistics systems. Only by utilising information network systems, can shippers construct state-of-the-art SCM systems. For example, cross docking systems for the advanced shipping notice (ASN) can be introduced, allowing deliveries to be expedited and a wider assortment of items to be offered, without maintaining large inventory stocks. Truck-matching systems using the Internet may improve loading efficiency, since it can decrease the number or empty trucks. Furthermore, IT may directly improve the situation of city logistics. Advanced IT systems such as intelligent transport systems (ITS) are expected to decrease traffic congestion. Hypotheses and their prospects. Most respondents agreed that IT will become widespread and will be utilized to streamline logistics activities. The development of IT systems relating to transactions, such as an open message exchange system across the industry (HI), logistics EDI using the Internet (H2) and ASN (H4) are viewed as likely developments. Furthermore, IT concerning transport and delivery such as cargo tracking systems using GPS (H6), computer-aided optimal dispatching
Future city logistics in Japan
271
systems (H5) and arrival prediction and detour directions systems (H7) are also viewed as likely. Many respondents indicated that they will try to take advantage of IT, as shown by the high correlation coefficient between evaluations and intended measures to cope (r = 0.88). There were no significant differences between the responses of shippers and those of carriers. Table 6 Hypotheses on information technology and results Hypotheses about Future (aroundyear 2010)
Shippers
. , ,
An open message exchange system across the industry will come into widespread use. Logistics EDI using the Internet will come into wide use even by small and medium companies. Truck-matching and warehousing-matching systems including shippers will come into wide use. Advanced shipping notice (ASN) will come into wide use. Computer aided optimal dispatching systems will come into wide use. Cargo tracking systems using GPS will come into wide use. Arrival prediction and detour directions systems using traffic information will come into wide use.
Respondents' Directions to Cope with the Hypotheses (*2) Shippers Carriers
Evaluations of Realisation (*1) Carriers
1.2
1.3
1.7
1.8
1.1
1.2
1.7
1.9
0.9
0.8
1.4
1.4
1.2
1.2
1.7
1.8
1.2
1.1
1.7
1.6
1.3
1.1
1.4
1.4
1.2
1.0
1.4
1.5
(Notes: Same as Table 3)
RECENT MEASURES TO BALANCE ENVIRONMENTAL CONCERNS WITH EFFICIENCY
Company policy papers on the environment hi the sections above, we have outlined the perspectives of shippers and carriers regarding various hypotheses about the future of city logistics. As the survey implies, one way to balance these considerations would be to introduce new logistics systems that employ the SCM and 3PL concepts, as well as IT capabilities. However, the JILS survey only asked respondents to offer their general opinions regarding certain hypotheses and did not generate any concrete examples of future trends in logistics. To make up for this deficiency, we have analyzed many policy papers on the environment that companies have published, to publicise the efforts that they are making to address environmental issues. Their efforts can be broadly categorised into those that address the issue of transport, those related to logistics centres, and those that focus on the use of packing and wrapping materials. We consider efforts related to transport issues to be most directly related to city logistics. In the discussion below, we focus on manufacturers, wholesalers and retailers of consumer goods with sales of more than one hundred billion yen.
272 Logistics systems for sustainable cities Among the manufacturers we examined, there were a total of 62 subjects, including 42 manufacturers of food products, 6 apparel manufacturers and 14 companies that produce goods for daily use (Table 7). Of the food manufacturers, 29 companies have published policy papers and one has promulgated a policy for environment. 27 companies have clearly established a transport policy that is designed to reduce the environmental impact. Manufacturers of beer and beverages are more proactive in this are, perhaps because their transport volume is far greater than that of other manufacturers. Of the six apparel manufacturers, only one has published a policy paper, whereas 12 of the 14 companies that manufacture products for daily use, have published policy papers and try to be environmentally responsible. Of the 20 wholesalers that deal with consumer goods, only two office supplies wholesalers have published policy papers on the environment. One reason why wholesalers may not be as eager to state their position is that they do not have to deal directly with consumers. Among the 68 retailers we examined, 26 have published policy papers on the environment, while 25 others have established environmentally sensitive transport policies and nine more have official statements of environmental policy. The attitudes towards environment tend to reflect their type of business; that is, supermarkets and convenience stores are more active in addressing environmental issues while speciality stores are less so. Overall, 66 companies have published a policy paper on the environment in every year, and have made efforts to plan, implement, monitor, and adjust their actions based on the policy paper. In following sections, we discuss in more detail the ways in which these companies have dealt with logistics, especially the issue of transportation. Table 7 How companies of consumer goods tackle with environmental issues Stated the Published policy environmentally sensitive papers on transport policy further environment 42 62 39 27 42 29 1 6 0
Number of Objects Manufacturers Foods Apparels Daily use Wholesalers Retailers Super markets Department stores Convenience Stores Speciality stores Total
14 20 68 24 12 8 24 150
12 2 26 13 5 8 0 70
Stated environmental policies only 1 1 0
2 25
0 2 9
12 5
5 1
8 0 66
0 3
12
12
Measures to reduce the environmental impact of transport Based on the policy papers on the environment published by many companies, it is clear that they regard measures to reduce the environmental impact of transport as an important issue. As for the methods used to make transport more environmentally sensitive, the most popular way is to establish collective delivery systems, with 63.6 percent of the companies studied having introduced such systems. Other measures that are popular include the introduction of cleaner trucks and rules to prevent drivers from idling their engines when stopped (Table 8).
Future city logistics in Japan
273
Reconsidering physical distribution systems. Twenty companies are reconsidering their physical distribution systems with the aim to decrease the volume of goods transported. For example, some manufacturers have relocated distribution centres, from remote areas to sites adjacent to their factories, in order to eliminate the need for transport from factory to distribution centre. Others have consolidated their distribution centres, to reduce transport volume. For example, Lion, a leading manufacturer of toiletries, reports that it was able to reduce the volume of goods transported by 30 percent over five years by integrating its distribution centres. Manufacturers with many factories tend to transport their goods over longer distances, because each factory produces different goods that are distributed throughout Japan. Some manufacturers are dividing their businesses into regional sectors, with the factories located in each region satisfying demand in that region. This helps reduce the distance of the distribution routes from each factory. For example, Ito En, a beverage manufacturer, divides its business operations into five regional sectors, while Takara Schuzo, a liquor manufacturer, operates six regional sectors. Retailers also have integrated distribution centres which used to be separated by product category into regional centres. However, since this tends to increase the distance from regional distribution centres to stores, some of them have built local distribution centres to decrease the total distance travelled. Ito-Yokado, a leading general merchandise retailer, has rebuilt a network of area distribution centres to serve their stores. As a result it has decreased the total transport distance per year from 440 million km to 254 million km. Table 8 Measures to reduce the environmental impact of transport Manufacturers of Manufacturers of goods for daily use foods and wholesalers Reconsidering physical distribution systems
Retailers
Total
11
40.7%
3
21.4%
6
24.0%
20
30.3%
Environmental transport management
7
25.9%
5
35.7%
7
28.0%
19
28.8%
Utilizing back hauls
3
11.1%
2
14.3%
0
0.0%
5
7.6% 63.6%
12
44.4%
5
35.7%
25
100.0%
42
Direct hauls
3
11.1%
0
0.0%
0
0.0%
3
4.5%
Changes in loading process Increasing load capacity with larger trucks Using rail and ship transport instead of trucks
7
25.9%
4
28.6%
0
0.0%
11
16.7%
5
18.5%
2
14.3%
2
8.0%
9
13.6%
Collective delivery systems
7
25.9%
6
42.9%
2
8.0%
15
22.7%
Introduction of cleaner trucks
15
55.6%
7
50.0%
17
68.0%
39
59.1%
Stop idling campaign
13
48.1%
8
57.1%
14
56.0%
35
53.0%
Total
27
100.0%
14
100.0%
25
100.0%
66 100.0%
Environmental transport management. Many companies conduct a thorough review of transport management with the aim of making their systems environmentally friendly as well as efficient. Suntory, a beverage manufacturer, reduced the number of trucks in use by 14 percent, by introducing a system of centralized information control on transport between warehouses. Snow Brand Milk Products reduced total transport distance by 30 percent, by introducing dispatching systems.
274 Logistics systems for sustainable cities Some companies have introduced on-board computerized management systems and global positioning systems (GPS) to manage the real-time status of their trucks. Some are controlling not only their own trucks, but also contracted trucks. Utilising bach hauls. Some companies are using truck-matching systems that match information on transport demand with information on vacant trucks especially in case of back hauling. Kewpie, a foods manufacturer, increased the load rates of its trucks from about 75 percent to more than 90 percent by introducing truck-matching systems via the Internet. Collective delivery systems. Many manufacturers and retailers have introduced collective delivery systems to decrease the number of trucks. Manufacturers, especially of processed foods, confectioneries and cosmetics, not only deliver their own products, but also carry goods produced by other manufacturers to wholesaler and retailer distribution centres. Ezaki Glico, a manufacturer of confectionaries, delivers 43 percent of their products using joint delivery systems. In the case of deliveries to general merchandise stores (GMS), vendors are asked to deliver their goods to distribution centres. The number of trucks making deliveries to the retail stores is reduced by consolidating loads at the distribution centre, rather than having each vendor make separate deliveries. Some GMS chains try to reduce the number of trucks further, by collecting goods from vendors on milk-run trucks. In the case of convenience stores, one important issue is to reduce the number of trucks making deliveries to each store. One example of the measures used to tackle this issue is the use of temperature-controlled collective delivery systems. Seven-Eleven Japan introduced collective delivery systems using four bands of temperature-controlled vehicle. This helped to reduce the number of trucks making deliveries to each store from 70 in 1974 to 10 in 2001 mainly by the effect of better load planning. Other convenience stores have introduced similar delivery systems (Figure 1).
Figure 1 Number of trucks making deliveries to convenience stores, per day (Source: Policy papers on the environment published by each company) Department stores have built distribution centres to concentrate shipments. However, many goods are still delivered directly to each store, especially in case of foods and major brandname products. These direct deliveries are not only inefficient but also help create traffic jams
Future city logistics in Japan
275
on the roads near stores. One good example is the Ikebukuro store operated by Seibu Department Stores. It covers an area of 63,470 square meters and generates 272,610 million yen in sales. The number of trucks making deliveries to this one store (one of the largest in Japan) was about 900 per day in fiscal year 1997. By introducing delivery agency systems, the company decreased this number to about 580 per day in 2001. About 75 percent of vendors utilize these systems, and the volume of direct delivery has decreased. Companies are also introducing collective delivery systems for home delivery from department stores. Mitsukoshi, Takashimaya, Matsuzakaya and the Daimaru jointly deliver about eight million parcels a year. Direct hauls from factories. Direct hauls from factories instead of from warehouses to wholesalers and retailers reduces the volume of transport by eliminating excessive loading and unloading. In particular, manufacturers of beverages and beer have tackled this issue with direct hauls. Asahi Breweries hauls 86.6 percent of its products directly from factories to wholesalers and retailers. As a result it has reduced the number of trucks. Changes in loading. Changes in the packing and wrapping materials used can reduce the volume of goods and increase the load rates as well as decreasing waste. In particular, manufacturers of foods and daily use goods are trying to adopt more environmentally sensitive methods. Some transport their goods by tank lorry instead of in paper bags, or have changed the method of stacking corrugated cartons to increase their load rates. Using such methods, Nissin Food Products increased the number of Cup Noodle items loaded on a truck with a ten tonne capacity from 16,800 to 25,200. Increasing load capacity with larger trucks. Increasing load capacity with larger trucks decreases the number of trucks, assuming that the amount of goods delivered is equal. Manufacturers have introduced larger trucks to transport goods to the distribution centres of wholesalers and retailers. Some food manufacturers have introduced trucks with capacity of 13 or 14 tonnes, instead of 10 tonne trucks. Using rail and ship transport instead of trucks. In the case of long-distance transport, for more than 500 km, some companies are trying to shift from trucks to rail and ship transport, which is more environmentally friendly. In particular, manufacturers are adopting this solution in the case of large-lot cargoes delivered from factories to warehouses. Ajinomoto, a manufacturer of food products, now transports 16.5 percent of all goods by rail container. As a result, it reduced the amount of CO2 exhaust emissions by 28,000 tonnes a year. Itoki, a manufacturer of office supplies, transports many of its products such as tables and partitions by rail. Mycal, a GMS chain, transports fresh vegetables such as lettuce from growing districts to distribution centres using temperaturecontrolled rail containers. The company transported 850 tonnes of vegetables in 170 container cars, from Hokkaido to Tokyo, and 1,150 tonnes of vegetables in 230 containers from Hokkaido to Osaka. Introducing cleaner trucks. 39 companies have introduced cleaner trucks to their fleets, including models that satisfy the latest regulations on emissions, compressed natural gas (CNG) trucks, and liquefied natural gas (LNG) trucks. However, few companies have asked their contracted carriers to introduce cleaner trucks.
276
Logistics systems for sustainable cities
Stop idling campaign. Many companies have introduced campaigns to discourage their drivers from idling at stops, sudden braking and sudden starts, to reduce the impact on the environment. Some companies have made a thorough effort to eliminate such practices, and monitor the actual situation. Seven-Eleven Japan and Ministop utilize on-board computerised management systems and GPS to monitor each truck, and determine whether they are idling, or making sudden stops and starts. Dispatchers can give advice to each driver individually, according to the monitors. Seven-Eleven Japan reported that it increased fuel efficiency of its trucks to 6.18 km per litre from 5.88 km per litre.
Volume of CO2 emissions In order to make a clear evaluation of environmental impacts, it is necessary to quantify the effects on environment. Of the 66 companies studied, 53 had a general grasp of the amount of energy consumed and CO2 generated by their activities. However, the level of quantification is still lacking in the production processes of most of manufacturers and in the stores of most of retailers. Only 24 companies can quantify the environmental impact of these activities, as well as the physical distribution process. One reason why few companies have considered this aspect seems to be that many companies commission carriers to do logistics activities. Thus, even if they quantify the physical distribution processes, they can calculate the impact only on the basis of conditions for their own trucks or those of their logistics subsidiaries. There are a few companies that calculated the environmental impact including trucks operated by their contractors. Almost all companies calculated the amounts based on their own transport activities to delivery points. Only Konica, a manufacturer of film, used a life cycle assessment (LCA) to calculate the amount for the entire delivery chain, including transport to final retailers. Although there are many limitations on the data, such as the points noted above, 17 companies were able to compare the amounts of CO2 exhaust generated in 2001 and 2000. 12 companies reported a decrease in CO2 emissions. The reason given for the increase in CO2 emissions for the other companies was that the volume transported had increased because of increased sales or changes to distribution systems (Table 9). Table 9 Changes in amount of CO2 exhaust produced by transport at each company 2001/2000 Category 2001/2000 Category 0.96 Supermarkets 0.85 Supermarkets 0.98 Supermarkets 1.23 Supermarkets
0.99 0.79 1.22 0.84
0.98 Supermarkets 0.98 Supermarkets
1.05 1.01
Manufacturer of foods
1.07 Department store
0.96
Manufacturer of film
0.99 Convenience store 0.95 Convenience store 0.93 (Source: Same as Figure 1)
Cigarette manufacture Brewer Manufacturer of beer and beverages Brewer Brewer Distiller
CONCLUDING REMARKS We have examined the views of shippers and carriers regarding the future prospects and likely direction of city logistics in Japan, based on the JILS survey. We also offered some concrete
Future city logistics in Japan
277
examples of how companies are making logistics systems both efficient and environmentally sensitive, based on their policy papers. The respondents believe that it is a matter of course that companies try to harmonise their logistics systems with the environment and cope with environmental problems. They also expect the government to tighten control over logistics activities to protect the environment. On the other hand, they think that the level of logistics services is still increasing and concepts such as SCM and 3PL will be introduced more widely. These opinions may be incompatible, with the respondents themselves are not convinced that the environmental impact of logistics will take precedence over the quality of logistics service. Environment-friendly logistics often entails higher costs, such as replacing diesel trucks with clean trucks, using rail transport instead of trucks, and so on. Furthermore, it can reduce the level of service in distribution. Therefore, one of the key challenges for city logistics will be the effort to maintain a balance between the environment and logistics service. Based on the policy papers on the environment published by many companies, there are some concrete examples to show the way. In the past, most companies aimed to make distribution systems more efficient, but now, they are trying to make them environmentally sensitive, as well. As the needs of consumers increase and diversify, companies are expected to maintain higher delivery service levels, including more frequent, small-lot deliveries, while decreasing the burden on the environment. To resolve these incompatible objectives, companies have introduced measures such as a review of distribution systems, precise transport management, use of back hauling, and so on. However, the impact of measures taken by single company is limited. Thus, some companies are also trying to introduce joint measures, such as collective delivery systems. When evaluating city logistics, it is important to understand the prospects of shippers and carriers, and the measures they will take to prepare for future challenges. We hope that the information presented here will be beneficial to academics and companies seeking to develop city logistics systems for the future. In closing, we would like to thank the JILS and the study group of comprehensive survey on logistics for giving us the chance to join the group and permission to use the precious data.
REFERENCES Bjorklund, M. (2001). Operational insufficiencies in creating an environmentally responsible distribution system, The 13th Annual NoFoMa Conference. Japan Institute of Logistics Systems (2002). Report on Comprehensive Survey on Logistics. Japan Physical Distribution Management Association (1984). Long-Term Vision on Physical Distribution Aiming at 21 Century. Romm, J. J. (1999). Cool Companies, Island Press, Covelo. Science and Technology Foresight Centre (2001). The 7th Technology Foresight Survey Future Technology in Japan. Yano, Y. and K. Hayashi (2002). The Shippers' and Carriers' Perspectives: Future Prospects of Logistics in Japan, Proceedings of the 7' International Symposium on Logistics, Monash Print Services, Caulfield.
This page is intentionally left blank
20
CITY LOGISTICS IN ITALY: A NATIONAL PROJECT
Simone Gragnani, Federtrasporto, Rome, Italy Gaetano Valenti, ENEA, Rome, Italy Maria Pia Valentini, ENEA, Rome, Italy
ABSTRACT One of the key priorities of the latest Transport and Logistics Strategic Initiative proposed by the Italian Ministry of Infrastructure and Transport (MIT), (Ministero delle Infrastrutture e dei Trasporti, 2001) is to improve the efficiency and sustainability of Italy's urban freight distribution network. In order to achieve this objective, the MIT has lent its support to a series of local projects currently underway in various Italian cities. These projects aim to co-ordinate the introduction of innovative policy measures, enhanced infrastructure, improved IT capabilities, and environmentally sound vehicles. This paper contains a broad overview of this major initiative, with sections dedicated to its background, aims and key challenges as well as to its on-going interaction with the project aimed at defining a National ITS architecture. Urban pilot projects are examined on a city by city basis with reference to the purpose, stated objectives and content of each. Finally the paper provides an analysis of measures designed to achieve economic efficiency and environmental benefits in the field of freight transport.
INTRODUCTION The Transport and Logistics Strategic Initiative {Piano Generate dei Trasporti e della Logistica - PGTL) was proposed by the Italian Ministry of Infrastructure and Transport and approved by the Italian Government in 2001. The Initiative expressed grave concerns about current trends in urban freight transport activities and the related negative environmental and social impact. In Italy, as in the other EU countries, urban freight transport patterns have changed substantially in recent years. These changes can be observed by the tendency towards a higher incidence of smaller deliveries. This reduces customer inventories but also involves
280 Logistics systems for sustainable cities higher transport costs and increased traffic volumes leading to greater congestion, road occupancy as well as problems related to parking, energy consumption and air pollution. The PGTL, aware of the urgency of the need for intervention, has come up with a series of measures designed to gain a better understanding of the basic characteristics and mechanisms of the urban freight transport system. It is also urging local authorities to adopt and implement urban traffic projects involving the application of innovative policy measures and city logistics schemes. These schemes aim to facilitate and improve the efficiency of urban freight operations and to minimise potential environmental and social damage. The PGTL has also proposed a number of immediate actions to define the National ITS architecture to pave the way for a faster and wider deployment of telematics technologies and services in the transport sector in order to increase safety and efficiency. The National ITS Architecture is aimed specifically at setting up a common technical and institutional framework so that public agencies and private organisations can fully deploy interoperable technological solutions and effective management strategies for the collection, processing and sharing of past and real-time transport data. The task of the National ITS Architecture is to define the functions that must be performed in order to implement a given service. To do this, it identifies the hardware and subsystems where these functions reside; the interfaces/information flows between various subsystems; and the communication requirements for such information flows. With a view to actively developing and testing the National ITS architecture, the Italian Ministry of Infrastructure and Transport has recently launched three pilot projects, one of which focuses on the application and integration of telematics technologies and services. This projects aim is to enable more efficient and cost-effective solutions in urban freight distribution and to promote innovation in city logistics to facilitate the expanding role of e-commerce solutions. The Urban Delivery and E-Commerce (UDEC) project is one in a series of major initiatives in city logistics and freight distribution. It is being promoted by a consortium comprised of 5 cities, Rome, Parma, Vicenza, Siena and Terni. The initiative involves the implementation of large-scale pilot projects. This project aims to integrate various ITS tools, promote the adoption of technologically advanced vehicles that run on cleaner fuels and introduce innovative policy measures and enhanced infrastructures. The main goals of the UDEC include the identification of viable strategies to improve freight and fleet efficiency, ensure customer satisfaction, promote environmentally friendly and energy saving policies. In particular, the project should lead to closer co-operation between suppliers, transport/distributors and logistic providers as well as savings in distribution costs and a reduction in traffic congestion levels and in pollutant emissions. As well as the 5 cities named above, the UDEC project consortium has numerous other partners, including Federtrasporto (the Italian national federation of passenger and goods transport firms), ENEA (the Italian Agency for new Technologies, Energy and Environment), CRF (the Research Centre of FIAT), TTS-Italia (the Italian Association of Telematics for Transport and Safety), local freight transport and logistics operators, consultancy companies (FIT Consulting, MEMEX, ITALIAMONDO) and others.
City logistics in Italy: A national project
281
This paper contains an overview of this major initiative, its background, objectives, key challenges and structure as well as its on-going interaction with the project aimed at defining a National ITS architecture. It also provides details on the purpose, expectations and content of each city pilot project, including an analysis of logistical measures designed to achieve economic efficiency and environmental benefits. Finally, this paper provides the framework for evaluating project results based on a common methodology for cross-city comparisons of impacts on transport systems, local economies as well as of energy saving and environmental policies.
ARTIST: A FRAMEWORK ARCHITECTURE FOR ITS IN ITALY The latest National Transport and Logistics Strategic Initiative (Ministero delle Infrastrutture e dei Trasporti, 2001) identifies technological innovation as being a priority in the future development of the Italian Transport System. Such an innovation should consist not only of environmentally friendly and safer vehicles but also include ITS. Following the example of the USA and France, the necessity of a framework Architecture for ITS was incorporated in the PGTL. Such an Architecture was defined as a strategy for identifying the functions, characteristics and interaction of all ITS components, i.e. technological systems, rules, users and service suppliers. It is a tool for achieving an open and competitive market as well as for ensuring consistency of information to end-users and interoperability between components even when produced by different manufacturers. At the end of 2002, the first version of the Italian Framework Architecture for Transport System was drawn up by the Italian Ministry of Infrastructure and Transport. (ARTISTTransport System Italian Telematics Architecture). The key characteristics of ARTIST are: > Compatibility with the European framework KAREN/FRAME > Compatibility with other aspects such as organisational architecture and computer navigation tools > Strict co-operation with pilot projects in the following thematic areas: 1. Hazardous freight; 2.Emergency calls; 3.Urban deliveries including e-commerce. ARTIST's goal is to provide tools to design innovative systems and in a subsequent phase, to incorporate the results of pilot projects. Nine "fields" have been singled out within ARTIST; they are: 1. Provide Electronic Payment Facilities 2. Provide Safety and Emergency Facilities 3. Manage Traffic 4. Manage public Transport Operations 5. Provide Advanced Driver Assistance Systems 6. Provide Traveller Journey Assistance 7. Provide Support for Law Enforcement 8. Manage Freight and Fleet Operations 9. Provide an Archive
282 Logistics systems for sustainable cities Within each of these fields ARTIST provides the logical architecture that depicts the input and output terminators of the entire ITS and defines the information flows into and out of the system. Moreover, the physical and organisational architectures are given in fields 1, 6 and 8 since they have been considered crucial to achieving sustainability in the transport system of the future. Of course, the UDEC project on city logistics is involved primarily in field 8. The general functional architecture of ITS for the "Manage Freight and Fleet Operation Management" (MFFO) field is illustrated below.
Figure 1 Main IT functions for Freight and Fleet Operation As can be noted, four "first level" functions have been identified in MFFO, i.e.: Manage Logistics and Freight (MLF); Manage Commercial Fleet (MCF); Manage Vehicle/Driver/Cargo/Equipment (MV/D/C/E); and Manage Multimodal Platform (MMP). MLF includes activities linked with the logistic chains from supplier to client; tasks related with intermodal transport have also been included within this main function. MCF covers tasks related to the planning and management of commercial fleets while MV/D/C/E comprises tasks carried out mainly by vehicle drivers supported by means of onboard tools. Finally MV/D/C/E includes the operations taking place within freight villages and bases. MFFO functions are mostly operated by two operator categories: haulage and logistics firms and platform managers. Although these two represent the main city logistics actors, other parties are also involved such as city administrators, retailers, and consumers. This means that implications in many other IT fields can be individuated such as electronic payment, traffic management, driver and journey
City logistics in Italy: A national project
283
assistance. These interconnections between telematics applications will be more fully illustrated in the following description of the UDEC project.
PROJECT DESCRIPTION
Outline and objectives The UDEC project constitutes an initial application and validation of ARTIST guidelines in the urban freight distribution sector. The projects mission is to identify new ways of addressing the issue of urban freight distribution and to determine the criteria and recommendations for a more effective deployment of telematics technologies in this sector. Special attention is also devoted to Ecommerce issues through the introduction of innovative logistics solutions aimed at facilitating the success of the companies involved. Several objectives are expected to be achieved following the application of both new logistics schemes and telematics technologies in urban freight distribution. The main objectives of the project are: > to reduce delivery costs; > to increase urban areas competitiveness; > to reduce traffic congestion levels in urban areas; > to reduce urban traffic emissions; > to support e-commerce development (both B2B and B2C); > to improve supplier-customer relations; > to increase the safety levels of persons and goods. The UDEC project will be set up in various cities in order to develop, test and compare a set of practical ways of improving urban freight distribution as well as to guarantee the usability of telematics architecture in various territorial and logistics organisation systems. The selected sites include: metropolitan areas (Rome), medium-sized towns (Parma, Terni and Vicenza) and small tourist-oriented towns characterised by a large historical urban centre which leads to specific requirements and high accessibility problems (Siena). These areas have been chosen because: > the different operational content of the individual projects - from a logistical point of view, provides an opportunity to compare systems in order to evaluate the most effective solutions and to verify the applicability of the architecture to the local situation; > the simultaneous start-up of the project in different areas enables the immediate verification of the effective ability of local systems to fully interoperate and cooperate with each other; > the proposed pilot areas, with their different characteristics, are representative of various urban situations. It is expected that the IT architecture for urban goods distribution will be designed, tested and verified by all interested parties, including individual customers, retailers and distributors. The
284 Logistics systems for sustainable cities ultimate aim of the project is to extend the architecture to a national level, so that other cities may compare different tested distribution systems. The project structure The project is comprised of 10 work packages. Only two of them will be developed locally at the test sites. To ensure greater coherence and consistency between different local projects as well a higher economy of scale, all activities concerning the design, impacts evaluation and validation of the architecture will be carried out centrally. A specific work package (WP) will explore possible financing instruments and mechanisms which could be used to support innovation in urban freight distribution, develop the site projects and export the projects to other urban areas. Given that a two-phase impact evaluation is envisaged (ex-ante and ex-post), a specific activity will be devoted to defining a suitable common methodology for all demonstration sites. A working group including representatives of stakeholders from various test sites will define the local IT architectures which must observe two constraints: >• the guidelines already set out in the National ITS architecture which ensure interoperability among sub-systems; > the current hardware and software equipment deployed in each test site, both in the public sector (traffic control systems) and in the private sector (fleet management, routing and communication with vehicles). After the first year of planning and simulation activities the test phase will be launched. The test activities involve cities where telematics systems for traffic control are already in place. As most of the distribution and logistic firms have their own communication and fleet management systems, it will be sufficient to adapt existing systems to ensure interoperability, by modifying existing hardware and developing new software. In these cases the project measures will concern infrastructures, on-board equipment (for vehicles) and dedicated software. In particular the following requirements should be met: > the electronic exchange of traffic and transport data and information among private and public operators; > the optimisation of routes and delivery times, according to the real traffic situation; > freight tracking through the entire distribution chain; > the minimisation of stoppages, empty miles and partial deliveries.
City logistics in Italy: A national project
285
Figure 2 Project overview Demonstration site projects Each local project plans to use one or more logistics platforms (or transit points) for receiving, stocking, storing and packing goods, as well as for scheduling and routing deliveries. Low-zero emissions vehicles shall be used for the last-mile haulage. The vehicles' size shall depend on the specific characteristics and requirements of each city. Rome Rome is the biggest metropolitan area in Italy both in terms of size (1,280 km ) and population (2,700,000 persons). There are 174,000 registered businesses located there with 670,000 employees, 150,000 of whom are engaged in the commercial sector. The test area covers 15 km2 and is in the historical centre of the city (enclosed by the ancient Aurelian Walls). There are approximately 150,000 residents, 15,000 firms and 137,000 employees working in the test site. A large part of the city centre is a Limited Traffic Zone (LTZ) where access is allowed to authorised vehicles only (residents and service vehicles) from 7.00 a.m. to 6.00 p.m. on working days. In particular for goods distribution, access is forbidden to vehicles exceeding 3.5 tons. Moreover, vehicles with a capacity of 1.5 to 3.5 tons cannot access the LTZ between 10 a.m. to 2 p.m. and from 4 p.m. to 8 p.m. So the time when vehicles can legally access the LTZ is actually quite short. Access is monitored by 22 gates equipped with a "telepass" system based on a transponder signal and video cameras connected to the Traffic Control Centre (IRIDE System). The number plate of a vehicle without a transponder is scanned by a video camera, fed into and verified by the Central Database and, if a violation has occurred, fined.
286
Logistics systems for sustainable cities
Freight traffic in central Rome accounts for just over 10% of total traffic but, due to the dimensions of the vehicles and frequent stoppages, freight vehicle volumes cause significant traffic congestion problems and greatly contribute to noise levels and pollutant emissions (STA and Comune di Roma, 2001). INTEGRAL, as the Rome project is known, involves a reorganisation of the freight distribution process in the central area, that could then be extended to the whole city. The new freight distribution system is based on the use of existing extra-urban private logistics platforms/warehouses. The freight will be gathered at these platforms and then transferred to a transit point close to the central area. This operation would be performed at night with medium/large optimally loaded vehicles, achieving greater efficiency than when smaller vehicles are used. At the transit point freight will be redistributed to several small vehicles, more suited to the historical centre streets that use cleaner fuel (GPL or CNG). Vehicles are operated with a high load rate and dedicated routing and scheduling software will be used to determine optimal routes for reducing mileage, travel time and pollutant emissions. Deliveries to shops could take place at any time during the day. The opportunity to derogate from the LTZ restrictions represents an incentive to use this new delivery system and compensates the higher costs incurred as a result of extra breaking up operations. Thanks to communication between vehicles and the Traffic Control Centre, scheduling and routing could be modified during the day to take real traffic conditions and unforeseen events (dynamic paths) into account. During the test phase the possible implementation of a remote control for the correct use of the loading bays will be verified (this is an important critical aspect of goods delivery in Rome). It should be possible to assign specific slots to vehicles. The project's main partners are: STA Spa (Municipality of Rome Traffic Agency) that manages existing traffic control systems and services and Federtrasporto, (the national association of transport firms). Other local partners include some of the most important distribution and logistics firms currently operating in the metropolitan area of Rome. These firms will place their logistics centres/warehouses at the project's disposal. Finally, the INTEGRAL project participates in the MEROPE project which is financed by the European Union as a part of the INTERREG HIB-MEDOCC Program. Terni Terni is a relatively small industrial town in Central Italy located near the Nera River Valley and the site of a dried lake in Italy's Umbrian Region. This region is characterised by the presence of a large number of small to medium towns distributed throughout its territory, which although relatively distant from one another, nonetheless enjoy strong economic links. In this sense, Umbria may be described as a "Regional Metropolis". In this context Terni is both a key national and a regional node. This in turn gives rise to a series of specific logistical issues. In fact Terni (110,000 inhabitants) represents the productive, commercial and transport core of the Region, with 12,600 firms, mostly in the chemical and steel sectors, employing 39,200 persons. There are 460 haulage firms located there as well as a large goods yard at the railway station which processes more than 1 Million tons/year. Most of the commercial, working and studying activities are concentrated within the area between the Orte-Falconara railway and the Nera River. This area corresponds with the
City logistics in Italy: A national project
287
nineteenth century quarter and the medieval centre, where a Limited Traffic Zone (LTZ) has been in place since the nineties. Freight transport represents about 12% of urban vehicular flows, a certain amount of which is made up of intra-regional crossing traffic. Terni is equipped with an Urban Traffic Control (UTC) System that has been designed to perform split, offset and cycle time optimisation for 20 signal controlled junctions under real traffic conditions. A supervisory system known as REGIT has also been implemented. The system uses data on traffic volumes to estimate the general traffic situation in the greater urban area as well as the level of noxious emissions, in order to set out possible control measures. The system is currently being improved to enable traffic and pollution forecasts in the short and medium period. The latest Urban Traffic Plan (2001) contains measures both for passenger and freight demand. Apart from some improvement in the viability of peripheral areas, the enlargement of several areas including a pedestrian zone and the electronic control of the LTZ is also planned, as well as the setting up of a freight logistics base. This base would be located beside the Orte-Terni Railway, at about 5 km from the Central Railway station and would receive freight traffic both from road and rail. The total amount of the investment for the logistics base is about 14 million Euro and a public and private partnership is now being established. A technological and operational scheme has been proposed within the project "FINESTRA", co-financed by the EU in order to carry out new solutions for SME and for production and haulage firms. At the same time, ENEA, under the auspices of the Ministry of the Environment, implemented a proposal for e-commerce deliveries by means of the logistics base and drop-points, in order to limit negative impacts on pollution and congestion. From an operational point of view, a new Urban Logistics Agency has also been planned. This would bring together the various actors in city logistics such as administrators, wholesalers and retailers, e-commerce operators, haulage and delivery operators, private citizens, consumer associations and city-terminal operators. The Technological Platform will represent the core of the logistic system in Terni. It will support logistic operators in planning deliveries and routes and optimising loading using analytical procedures fed with incoming data from remote traffic surveillance and control systems, clients and operators. The following scheme illustrates the functions and information flows to/from the technological platform in Terni.
288 Logistics systems for sustainable cities
Figure 3 Functions and information flows to/from the technological platform Within the UDEC Project for a national architecture, the Terni project will focus on testing the links between the technological platform of the logistics base and the traffic control system. This could prove to be of vital importance especially in a wider urban context where duty vehicle routing is strongly affected by traffic conditions. By means of "dynamic simulation", a real time control policy could be adopted in the future, making telematics more and more crucial. Siena The Siena Municipality project has the following features: • two terminals, one for consumable goods and the other one for non-consumable goods both serving the historical centre of the ancient town which is a Limited Traffic Zone; • a public-private partnership for the construction and management of the two bases; • a technological platform for the management of commercial transactions and vehicle routing; • automatic access control system already implemented around the LTZ, verifying both in-coming and out-going traffic flows so that the monitoring of stoppages and, to some extent, vehicle routes is possible; • smart card to access parking and riding facilities. An environmental impact assessment methodology has been developed for this project (Valentini et al, 2001). Siena's role within the UDEC Project will be the checking of automatic controls, reconnoitring and payment systems for freight vehicles and drivers linked with the logistic bases. In other words, Siena will check how, and how efficiently, controls and facilities are able to ease the introduction of a new organisation in urban goods delivery.
City logistics in Italy: A national project
289
Parma The city of Parma is located in the centre of Northern Italy, mid-way between Milan and Bologna. With a population of 170,000 (more than 40% of the provincial area), the city is situated in the fertile valley of the Po River in the heart of one of the richest Italian regions, the Emilia-Romagna. The province of Parma has a strong industrial sector with a high percentage of small sized companies and a thriving commercial sector. The retail sector plays a key role in Parma's economy and is dominated by small-scale, family owned outlets (many of which cover an area not exceeding 250 square meters). About 26 per cent of the 7,000 retail outlets operate in the foodstuffs sector. The city of Parma has long ago started projects aimed at deploying telematics technologies for improving traffic and mobility management. More specifically it has implemented a supervisory system that allows operators to continuously monitor traffic performance and air quality in the city area. The supervisor combines and integrates information and data coming from the dynamic traffic signal control system, the traffic monitoring sensors placed on the main streets, the video-surveillance devices installed in some junctions and the local air quality monitoring stations. Currently the city of Parma is completing further ITS projects to facilitate payment for parking in a central area of the city and to limit access by road traffic to the historic centre. The access control system will use an electronic tag which permits authorised vehicles to pass an electronic gate without stopping, while an on-line video enforcement is planned to allow the system to automatically record the passage of any unauthorised vehicle and, in this case, issue a violation notice. The city-logistics pilot project in Parma is intended to strengthen co-operation between the many operators responsible for carrying out urban freight distribution with a view to improving the efficiency of freight distribution and to reduce its harmful impacts. The project is expected to involve the whole city area and is articulated in two consecutive phases. The first phase envisages, as a preliminary test site, the historic centre where it is estimated that deliveries account for approximately 40% of the total amount of delivery operations in the urban area and where the delivery density per square kilometre is seven times higher than in the suburbs. In the second phase, it is proposed that the test site be extended to include the entire city area. The project plans to set up four city transit points located in the suburbs near the city ring road where goods can be received, stored, consolidated and sent to the final customers. The transit points will be managed by logistics operators with qualifications and competence assured through an accreditation system but without exclusive rights to use these facilities. The freight distribution service is set to be performed with low-zero emissions vehicles in order to guarantee deliveries to the historic centre even during the hours when conventionally fuelled vehicles are banned. A fleet of 10 low/zero emissions vehicles shall be operational during the first phase of the project. In order to solve the "last mile" problem that prevents the local growth of the e-commerce market in the Business to Consumer segment, the project also intends to identify a network of suitable drop-off points in the urban area, such as post offices and service stations. The adoption of new logistics schemes in Parma will be based on the application of advanced telematics solutions enabling operators to manage and perform a set of operations in a costeffective and efficient way. In particular the project involves the application and integration of a set of telematics technologies to provide a basic structure for the electronic data processing of
290 Logistics systems for sustainable cities logistics information chains from source to destination and for the integrated planning, processing, monitoring and controlling of transport flows from the supplier to the final customer. The project also includes the deployment of other complementary telematics technologies to handle fleet operations including the scheduling and specification of driver tasks and vehicle maintenance and to manage freight and fleet operations that are on-board the freight vehicle. The technological platform supporting the set of logistics/transport operations is planned to interoperate with the other urban ITS tools used to manage and control traffic, thus ensuring that logistics and transport operators can make use of real-time traffic information to better handle their resources. The main expected benefits of the new approach to city logistics in Parma include a reduction in current operating costs and externalities such as delays, accidents, energy, noise and pollutant emissions associated with urban freight transport. To this end, the project envisages further actions aimed at promoting broader co-operation in the use of transit points by local actors in the freight distribution process; optimising delivery/collection processes in relation to either the route and/or the vehicle capacity and the established loading/unloading times, reducing the number of "empty miles" and encouraging a higher utilisation of clean vehicles. Vicenza The city of Vicenza is a middle-sized city in the north-east of Italy with slightly more than 100,000 inhabitants (about 780,000 inhabitants live in the provincial area). It is situated 40km inland from Venice. Vicenza is at the heart of one of Italy's biggest export regions, with a large number of small firms producing goods and services for international markets. Currently the city of Vicenza deploys a real-time computerised traffic signal control system that monitors traffic conditions and selects appropriate signal timing (control) strategies. There are also 20 Variable Message Signs (VMS) installed along the main access roads of the network displaying real-time traffic information to road users. On-going ITS projects in the city are expected to lead to the development of an automatic access control system to the limited traffic zone and the integration of the current set of traffic monitoring devices placed on the road networks with a video camera system. Recently the city of Vicenza launched a project aimed at overhauling the urban freight transport system as part of a series of measures planned to address growing traffic congestion, to improve air quality and to increase the competitiveness of the urban area. The project is expected to favour more sustainable patterns in urban freight transport through the setting up of highly organised logistical chains within the city involving craft, trade and service sectors and the introduction of clean vehicles that minimise environmental damage. The project will involve a large part of the urban area where about 5,000 commercial activities (mainly in the clothing and refreshment sectors) are based. A recent survey shows that the annual number of deliveries to this area amounts to 671,000, corresponding to 47,000 tonnes of freight. The project aims to build a logistics centre in the outskirts of Vicenza managed by a company (Vicenza City Logistic - VCL) which shall be controlled, for the most part, by public bodies and with stakeholders in private associations and in the trade, industry and transport sectors. Advanced telematics applications will be set up to support a high number of functions resulting
City logistics in Italy: A national project
291
from the complexity of the supply chain and from the wide variety of final customers. The telematics applications will be structured according to different functional levels. This will involve diverse operations such as: 1. the planning, monitoring and controlling of the flows of goods from the supplier to the final customer 2. a variety of warehousing tasks such as receiving, stocking, storing, picking and packing 3. the processes of vehicle routing and scheduling 4. administrative tasks such as accounting, management reporting and statistics The logistics centre will also manage the collection of packaging and products at the end of their life span, for subsequent recovery or disposal. The project will set up co-ordinated freight distribution patterns able to increase the current low load factor and thus to cut vehicle mileage in the urban area. The resulting reduction in the number of freight vehicle-kilometres, together with a set of policy measures planned to facilitate urban freight operations (such as improved on-street loading/parking facilities for freight vehicles and improved traffic/roadwork information) are expected to significantly lessen network congestion and local air and noise pollution. According to estimates, the main expected quantitative benefits of the introduction of the new city logistics schemes will be a reduction by 20-30 % of current pollutant levels and 40-50% of fuel consumption associated with the distribution activities. Moreover, a reduction of approximately 20% of the current operational transport costs is expected to be achieved.
Environmental impact assessment Environmental issues are a decisive factor in the adoption of city logistics schemes. That is why a certain amount of the project's resources have been devoted to assessing the environmental impacts of the new logistics schemes planned in the five pilot sites. Such an assessment will be performed by means of a common methodology defined by ENEA and organised into two phases: ex-ante and ex-post evaluation. Ex-ante evaluation requires a simulation of the changing attitudes, new rules and opportunities which the new system is expected to bring about. In principle it will be necessary to estimate the amount of logistics operators expected to adhere to the new organisation and, assuming optimal operational modalities, to verify the reduction in total distance covered and changes in vehicular technology and operational conditions, such as mean speed and maintenance. Analyses will be carried out using the available data on duty vehicles operating in each test area as well as on shops, stores and any commercial or productive activity within the area. In some cases it could be necessary to integrate such data with specific surveys. It is feared that major variations in data availability could be the main obstacle to outlining a common ex-ante evaluation methodology. In any case, a set of indicators shall be defined, such as mean vehicular loading and filling times, total and mean distance covered, unit and total
292
Logistics systems for sustainable cities
energy consumption and pollutant emissions, so that by using a bottom-up and/or top-down methodology, a case by case comparison will be possible. Also an economic assessment will be carried out comparing new and old costs, both internally and externally. Clearly, ex-post evaluation shall be conducted once the project has been implemented. The aim is both to monitor the project's progress and to track the changes in the efficiency and effectiveness of delivery schemes. The success of single initiatives will depend on organisational and technological performances. Monitoring will be used to verify both of them.
CONCLUSIONS The UDEC project is currently a major initiative in the city logistics and freight distribution area and involves 5 cities (Rome, Terni, Siena, Parma and Vicenza). The project aims at promoting the development of large-scale demonstration projects integrating aspects of ITS tools, new technology vehicles with cleaner fuels, innovative policy measures and infrastructure facilities. The purpose of this project is twofold: first, to test viable strategies for freight and fleet efficiency, customer satisfaction, energy saving measures and environmental improvement, and second, to develop and validate ITS equipment and services for Freight and Fleet Operation Management within the on-going National ITS architecture development project which is intended to provide a common framework for planning, defining and integrating ITS technologies. It is envisaged that the UDEC project will promote closer co-operation between suppliers, transport/distributors and logistics providers, enhancing global operational efficiency as well as reducing traffic congestion levels and air pollution. In particular the main expected benefits of the adoption of the new city logistics projects range from a 20% to 50% reduction in noxious emissions and energy consumption, depending on the involvement of suppliers and transport operators in each urban context and on the use of clean vehicles for freight distribution operations. Moreover, rough estimates do not envisage significant variations in the cost base of current and future scenarios. In fact, the higher consolidation costs will be largely offset by lower operational costs associated with a higher loading rate of the vehicles.
REFERENCES ADEME (2002). French National Research Program on Urban Freight: tools to learn, understand and apply, Presentation at BESTUF Conference (26th and 27th June 2002), Paris. Browne, M. (2001). E-commerce and Urban Transport, CEMT Seminar: The impact of ecommerce on transport (5th and 6th June 2001), Paris. Carrara, M. and Monticelli, M. (2000). Long Term Forecasts of Urban Freight Distribution in Europe, CSST.
City logistics in Italy: A national project
293
MEMEX (2002). FINESTRA Project Del. 5 - Modello di riferimento ed architettura tecnologica Deliverable 5tudioper la mobilita delle, Livomo. Ministero delle Infrastrutture e dei Trasporti (2001). Piano Generate dei Trasporti e della Logistica, Roma. Ministero delle Infrastrutture e dei Trasporti (2002). ARTIST (Architettura Telematica Italiana per il Sistema dei Trasporti), Versione 1, Roma. Lacquaniti, P., M.P. Valentini (2000). - Progetto pilota per la distribuzione delle merci nel centro storico di Siena: valutazione degli impatti energetici ed ambientali- ENEA Casaccia Roma. Romanazzo, M., G. Valenti and M.P. Valentini (1999).- Progetto pilota per la distribuzione delle merci nel centro storico di Siena: linee progettuali- ENEA Casaccia Roma. STA and Comune di Roma (2001). Studio per la mobilita delle merci nel centro storico di Roma, Roma. Valentini, M.P., P. Lacquaniti and G. Valenti, (2001). Methodology and results of a study on logistic schemes in Siena, City Logistics II, Proceedings 2nd International Conference on City Logistics (E. Taniguchi and R.G. Thompson, Eds), Okinawa, Institute for Systems Science, Kyoto.
This page is intentionally left blank
21
DEVELOPMENTS IN URBAN DISTRIBUTION IN LONDON Mem Baybars,Service Development Adviser, Transportfor London (TfL) Street Management, London, UK Michael Browne, Exel Professor of Logistics, Transport Studies Group, University of Westminster, London, UK
ABSTRACT The paper presents a review of distribution developments in London. It starts by discussing the importance of freight transport activity in London and then outlines some of the major policyrelated changes that have resulted from the new political structure resulting from the election of a Mayor for London in 2000. The importance of the Mayor's Transport Strategy and in particular the creation and role of the London Sustainable Distribution Partnership (LSDP) are considered. Other recent initiatives such as the congestion charging scheme (operating from 17 February 2003) and the feasibility studies for low emissions zones are also explored.
BACKGROUND TO NEW GOVERNANCE IN LONDON The Greater London Authority (GLA) was created by the Greater London Authority Act 1999. The GLA covers the area of 33 London boroughs, including the Corporation of London. It is made up of a directly elected executive Mayor and a separately elected Assembly, which primarily exercises scrutiny functions. The first mayoral elections took place in May 2000 and Ken Livingstone was elected the first Mayor of London. He took office in July 2000. The GLA Act identified the principal purposes of the GLA as promoting London's economic and social development, wealth creation, and environmental improvement. The Act requires the development of eight statutory Mayoral strategies covering: (a) Transport (July 2001) (b) Spatial development (draft London Plan June 2002) (c) Economic development (July 2001)
296
Logistics systems for sustainable cities
(d) Air Quality (September 2002) (e) Ambient noise (Draft July 2002) (f) Biodiversity (July 2002) (g) Municipal waste (September 2002) (h) Culture (February 2003) The Mayor has also produced an Energy Strategy. He is an enthusiastic supporter of hydrogen power and has set up a London Hydrogen Partnership to promote the development and use of hydrogen fuel cell technology in London. The GLA is comprised of four functional bodies: London Development Agency, London Fire and Emergency Planning Authority, Metropolitan Police Authority and Transport for London (TfL). Transport for London is responsible for the London Underground, the Strategic Road Network, Docklands Light Railway, London Bus services, Traffic Control Systems, and London River services. It was responsible for developing and implementing the central London Congestion Charge scheme and runs the London Traffic Control Centre which controls all traffic signals in the Capital.
THE MAYOR'S TRANSPORT STRATEGY The fundamental policy direction of the Mayor's Transport Strategy (as with the Mayor's other strategies), is to support investment in public infrastructure to accommodate London's growing population and economic activity. London's population is set to grow to 8.1million by 2016, from 7.4million today. London's GDP represents around 20% of the UK's entire output, and is bigger than many national economies. The City of London alone generates over £30 billion a year for the UK economy. It turns over $600 billion in foreign exchange transactions each day: a third of the world total, and twice as much as the USA. London is a World City. The Mayor's Transport Strategy aims to respond to the challenges that a World City brings - the needs of the citizens; growth in population; growth in the economic sector and the environment; reducing traffic congestion and improving public transport. The Transport Strategy is closely linked to other strategies produced by the Mayor, but in particular, to the Spatial Development Strategy, named as the London Plan, and the Economic Development Strategy. The draft Transport Strategy underwent a detailed public consultation. The Central London Congestion Charging scheme was a major feature of the draft strategy and it was the subject of additional consultations, culminating with formal consultation on Traffic Orders, from 23 July to 28 September 2001. This aspect of the mayor's work, producing new strategies, is probably unique in Europe. Since the GLA was set up as a new organisation, the legislators placed stringent requirements on the Mayor and GLA, to produce eight strategies covering all issues affecting a World City in the 21st century. It was stipulated that these strategies be developed in close consultation with all stakeholders including the 33 London local authorities, business sector and numerous interest groups. The linkages between the strategies are also important and some, inevitably could give rise to conflicts which need to be resolved. For instance, in the context of this paper, there
Developments in urban distribution in London
297
could be conflicts between the transport policies for performing more goods deliveries at night, thus helping to reduce day time congestion, and the Ambient Noise Strategy which would seek to reduce noise nuisance during the sensitive hours of the night. There will be other examples and these will become apparent as the strategies are put into practice. But probably the uniqueness of the London situation compared with other large cities in Europe, is that these very topical strategies are being produced from scratch and the synergies are being explored, conflicts identified and dealt with. The creation of the GLA has created the conditions for London to progress towards an integrated transport, development and environmental strategy in the first quarter of the 21 s t century. The Strategy takes reducing congestion and the increase in the overall capacity of the transport system as key priorities. Another key priority relates to the movement of freight: "Making the distribution of goods and services in London more reliable, sustainable and efficient, whilst minimising negative environmental effects." It is the policies, strategy and actions on freight that this paper will concentrate on.
THE MAYOR'S TRANSPORT STRATEGY AND FREIGHT Achieving an efficient and sustainable distribution system for goods and services is one of the greatest challenges facing London. In the recent past, freight has occupied something of a 'Cinderella' position in transport policy. The Government's Sustainable Distribution Strategy published in 1998 provides an overall policy framework, emphasising the twin goals of increased efficiency and reduced environmental impacts. The Mayor's Transport Strategy builds on these national objectives in London, and grasps the new opportunities offered by the creation of the GLA, as mentioned above, in giving leadership within the context of a holistic, partnership driven approach. In respect to distribution and servicing, the Transport Strategy seeks to: (a) ensure that London's transport networks allow for the efficient and reliable handling and distribution of freight and the provision of servicing in order to support London's economy; (b) minimise the adverse environmental impact of freight transport and servicing in London; (c) minimise the impact of congestion on the carriage of goods and provision of servicing; (d) foster a progressive shift of freight from road to more sustainable modes such as rail and water, where this is economical and practicable. One of the main mechanisms that the Mayor has put in place in order to bring about the above aims was the creation of the London Sustainable Distribution Partnership (LSDP). The LSDP, which can be considered as a Logistics Platform, was established in early 2002. The objectives of the LSDP are encapsulated in the Mayor's Transport Strategy. Policy 4K.1 states: "The Mayor and Transport for London will work with the London boroughs, business and freight, distribution and servicing industries, and other relevant organisations to ensure needs of business and Londoners for the movement of goods (including waste) and services met, whilst minimising congestion and environmental impacts in accordance with objectives of the Mayor's Transport, Air Quality, Waste and Noise Strategies."
the the are the
These objectives are being achieved through the joint working of the partners under the umbrella of the LSDP. The members of the LSDP are drawn from a diverse group but all with
298
Logistics systems for sustainable cities
an interest in freight and its commercial and environmental impacts. The value of such a diverse group is to bring expertise, knowledge and examples of good practice to deliveries, distribution and servicing in London. Members of the LSDP include: The Greater London Authority; London Development Agency; The Department for Transport (DfT); Freight Transport Association; Association of London Government; Road Haulage Association; British Retail Consortium; Association of Convenience Stores; Association of the International Express Couriers; Rail Freight Group; London Forum; Port of London Authority; London First; London Chamber of Commerce and Industry; Transport 2000; Confederation of British Industry (CBI); Strategic Rail Authority and the University of Westminster. LSDP has identified the key issues affecting sustainable distribution in London and is encouraging the introduction of trial schemes and championing innovative solutions. The current work areas of LSDP are: (i) road based distribution and delivery issues; (ii) rail freight development; (iii) the use of London's waterways. In addition, the Transport Strategy attaches considerable importance to the setting up and operation of sub-regional Freight Quality Partnerships (FQPs) and some sub-regions are already active in this area. LSDP is providing guidance to the fledgling FQPs.
FREIGHT TRANSPORT ACTIVITY IN LONDON NOW In 2000, 123 million tonnes of road freight lifted by vehicles with gross weights of over 3.5 tonnes had its origin or destination in Greater London, 4 per cent more than in 1999. For 50 million tonnes, both the origin and destination were in London. Total freight lifted by road in Britain was 2 per cent more in 2000 than 1999. However, the freight lifted in London in 2000 is approximately 17% below the 1990 level, reflecting changes in the nature of London's economy. The freight lifted in London represented approximately 8% of the total freight lifted in Britain by weight in 2000. About 80 per cent of the UK's air freight passes through the London area airports, with Heathrow handling 55 per cent of all UK air cargo and servicing a major concentration of freight shipping and forwarding activity in the Heathrow area. In 2000 about 1.8 million tonnes of airfreight was handled at London airports (almost double the volume for 1990). Heathrow handles the majority of London air freight. Sea-going freight traffic through the Port of London, which is the largest port in the UK, declined between 1990 and 1992, then increased to 57.3 million tonnes in 1998 before falling to 47.9 million tonnes in 2000. Over the last decade, freight traffic through the Port of London has fallen by 18 per cent. Internal freight traffic on the Thames has fluctuated over the last decade, in 1999 it was 15 per cent below the 1990 level (1.7 million tonnes compared with 2.0 million tonnes). Figure 1 shows the total volume of freight, in terms of tonnes moved, between 1990 and 2000. This illustrates that road is the dominant mode for goods movement in London. If Port of London traffic is excluded from the calculation, road accounted for approximately 94% of all freight lifted in London in 2000. Figure 1 only includes rail freight data up to 1994. Annual rail freight data has not been available since then. However, in 2000 it was calculated that rail freight accounted for approximately 4% of total freight lifted in London.
Developments in urban distribution in London
299
Figure 1 Freight moved in London by mode Throughout the 1990s, the total volume of motor traffic in Greater London has remained steady at around 29 to 30 billion vehicle-kilometres. Around 38 per cent of traffic is on principal roads, 32 per cent on minor roads and 30 per cent on trunk roads. Cars comprise over 80 per cent of the total vehicle flow on major roads in London, while commercial vehicles for freight and servicing account for approximately 14 per cent. Many freight and service movements now use the M25 rather than passing through London, including much of the road access to the east coast ports. The 1990s have also seen a shift in freight traffic from medium sized goods vehicles to light vans. Figures 3, 4 and 5 show the numbers of medium and heavy goods vehicles crossing into outer, inner and central London to be about stable or falling with the main growth being in light goods (van) traffic, especially in inner and outer London. Figure 2 shows the location of the boundary and cordon points. In the absence of policy change, total goods and service vehicle traffic in London is projected to rise by over ten per cent by 2011, with light vans likely to continue to be the fastest growing category. Moreover, the trend is for the number and variety of distribution services to increase with e-commerce and home deliveries, the 24-hour city and increasing consumer sophistication.
300
Logistics systems for sustainable cities
Figure 2 Location of London road traffic cordons
Figure 4 Inner London cordon daily crossings - freight vehicles: 24 hour flows
*&"* 3 G L A b o u n d a r y d a i l y c r o s s i n S s - freight vehicles: 24 hour flows
Figure 5 Central London cordon daily crossings - freight vehicles: 24 hour flows
Even with greatly increased use of other modes, the vast majority of goods transported in London will continue to be road based. Road space in London is at a premium and needs better management. There are no easy solutions to the many potential conflicts. It is within this context that opportunities for improving freight movements and servicing must be found.
NEW DEVELOPMENTS AND INITIATIVES There are a number of new major developments resulting from the Mayor's Transport Strategy that have implications for freight transport in London, such as: (a) Congestion charging scheme in central London; (b) Low Emission Zone (LEZ) feasibility study; (c) Review of the London Lorry Control Scheme. These are covered in more detail in the subsequent sections. In addition, TfL, LSDP, FQPs, the boroughs and the Strategic Rail Authority are tasked with carrying out reviews and developing initiatives in the following areas, which are described below. Allocation of road space - in terms of moving traffic, TfL is examining the scope for Priority Lanes (non-car lanes) and their implications for other road users, primarily cyclists. These may be appropriate on major corridors where bus flows alone do not justify a bus lane, or in
Developments in urban distribution in London
301
particular circumstances where high heavy goods vehicle flows are present. TfL is conducting this work in liaison with the Freight Transport Association, (FTA). Provision for Loading and unloading - there is work to be done to provide for the legitimate loading requirements of business, in particular, in bus lanes. On the Strategic Road Network, there are designated loading boxes which can be used for a 20-minute period at a time. However, there are problems with illegal parking. With the increasing use of enforcement cameras it may be possible to start a more intelligent enforcement regime. A study is underway to investigate the issues surrounding this initiative. Servicing - the servicing requirements of commercial (and residential) premises have grown substantially in recent years. There is, however, no regulatory definition of servicing, nor means of distinguishing vehicles involved in servicing. This makes provision of parking spaces specifically for servicing difficult. This issue is being considered and TfL is looking to identify a local pilot scheme to explore the options with the boroughs. New means of delivery - the aim is to see distribution become more efficient in terms of the amount of travel involved in deliveries, the adverse environmental impacts associated with travel and the utilisation (load factor) achieved. Initiatives that could be considered include: consolidation of loads and numbers of deliveries; changes to delivery hours - to extend the times during which deliveries can be made by addressing regulatory and commercial restrictions that currently prevent this; use of vehicles better suited to operation in dense urban environments, in terms of size and pollution. A nationwide initiative which has the backing of the Government is the lifting of restrictions on deliveries after 21.00 hours. The proposals for a pilot scheme in London will involve the operator and the store to use vehicles that run on cleaner fuels such as CNG, and employ methods which will reduce noise at the point of delivery. Information technology - this has the potential for delivering significant environmental benefits to the distribution sector in London. The LSDP and FQPs are seeking to encourage schemes promoting new approaches in these areas. Road freight and the environment - encouragement of operators to: make maximum use of Government programmes to bring about a switch to cleaner and quieter vehicle technologies; implement quieter freight, distribution and waste operations and practices; and implement high standards of vehicle maintenance, and considerate and economical driving. Rail freight - the Strategic Rail Authority (SRA, 2001) forecasts indicate that approximately 20 per cent of the predicted growth in rail freight nationally could arise in London, with the majority of this being non-bulk items transferred from other modes. However, UK rail accidents have affected confidence in rail freight. To achieve this growth in London's rail freight would require substantial increases in handling facilities. It indicates that three or four intermodal freight handling facilities would be required. Feasibility studies are also being carried out into using some central London rail terminals for freight distribution at night. Water-borne freight - the River Thames provides significant opportunities for sustainable freight access into the heart of the Capital. The Thames is particularly suited to the transport of bulk materials, such as waste and aggregates. The movement of waste by river is largely dependent on the continued availability of waste disposal (landfill or incineration) on the Thames. GLA is initiating a collaborative approach across London, which focuses in particular
302
Logistics systems for sustainable cities
on protecting existing facilities and supporting water-borne freight movement through the planning regime. Airfreight - There may be scope for some decentralisation of Heathrow's air freight activities to other sites that can be served by rail or other low emission vehicles. More use of low emission vehicles between depots in the Heathrow area, and between depots and aircraft will be encouraged.
CONGESTION CHARGING IN CENTRAL LONDON Transport for London introduced the central London Congestion Charge on the 17th February 2003. This is an area-based scheme, and not a cordon charge, for vehicles which are in the charge area between 7:00 and 18:30hrs. The charge area is approximately 5 km from east to west and 4 km from north to south. The charge is £5 (7 Euro). Goods and service vehicles working in central London will be subject to this charge. Research carried out by TfL prior to the introduction of the charge indicated that congestion charging would result in substantial decreases in traffic: (a) Inside the zone: (i) traffic would be reduced by 10 - 15% (ii) queues would be reduced by 20 - 30% (iii) traffic speeds would be increased by 10 - 15% (b) Outside the zone: (i) traffic may increase on orbital routes by up to 5% (ii) traffic would be reduced on radial routes by 5 - 10% (iii) overall reduction in traffic by 1 - 2% TfL considered that the forecast reduction of 10-15% in traffic levels would be sufficient to ease congestion and that the essential traffic such as buses and freight and service industry would benefit from shorter and more reliable journey times.
Developments in urban distribution in London
303
Figure 6 Central London Congestion Charge Area It was originally proposed that goods vehicles should pay £15 per day. The freight industry was critical of this charge and following the consultation on Congestion Charging, the charge was reduced to £5 per day for all vehicles. However, not all goods and service vehicles are subjected to the charge. The Mayor is also using the charge to encourage the vehicles which serve London to be "cleaner". Heavy Goods Vehicles which meet the Euro III standards are eligible for 100% discount. Lighter goods vehicles are benefiting from similar discount, if they are 40% cleaner than Euro IV standards in nitrogen oxide and hydrocarbon emissions. Electric vehicles also receive the same discount. Motorbikes, mopeds and cycles are exempt from the charge and do not have to be registered. The consultation on the payment of the congestion charge with the Freight Transport Association and other industry representatives have resulted in the creation of the Fleet Scheme. This scheme entitles operators with more than 25 vehicles to enter a special payment arrangement. For HGVs this involves the operator having a monthly account with TfL, depositing the required amount monthly which is then automatically decremented when cameras register operator's vehicles. Congestion charging is being enforced by a network of about 700 automatic numberplate recognition cameras (ANPR), located across about 250 sites (including all entry and exit points to the zone). These are supplemented on the Inner Ring Road and main radial approaches by 70 monitoring-only cameras. These cameras provide a new and potentially diverse set of data describing traffic flows in and around the zone, allowing (subject to Data Protection measures) vehicles to be categorised and their travel behaviour measured. Since the various types of freight/delivery vehicles can be separately identified, data from these cameras can potentially be used to significantly improve our understanding of the characteristics and behaviour of freight traffic in and around central London. Clearly, reductions in traffic could lead to greater reliability in trip times for goods and service vehicles. Increased reliability would off-set some or all of the additional costs but, prior to the introduction of the scheme, there remained some uncertainty about the likely impacts. In addition, while it was argued that traffic would fall in the congestion charging area it was also claimed that congestion would be worse around the edge of the zone. This in turn would reduce
304
Logistics systems for sustainable cities
the level of benefits to be expected from more reliable delivery and service trips in the central area. Now that the Congestion Charge has been in operation for almost six months, it is possible to test some of the predicted outcomes. Comprehensive monitoring surveys are being carried out and they will be able to demonstrate the impacts of the charge on commercial activity as well as the traffic movements within and around the zone. The full survey results will not be available until later in the year but the initial results based on the first three-month surveys indicate that (Transport for London, 2003): (a) A full survey of all vehicles coming into the charging zone during late February and March 2003 shows a reduction of around 20% compared to the equivalent time one year ago; (b) Traffic levels inside the zone (taken from sample counts immediately after the scheme's introduction) were 16% lower than equivalent counts taken at the same time in 2002; (c) Average traffic speeds across the charging day have increased from 13 km per hour in June 2002, and 15km per hour just prior to the scheme, to 17 km per hour in June 2003; (d) Journey time to central London from the rest of Greater London has reduced by 14% since the scheme was introduced; (e) The net volume of diverted traffic has been relatively small; (f) Bus patronage is up by 14% since the start of the scheme. London First, a prominent business lobby group, carried out a survey of its members a month after the introduction of the scheme. The survey showed strong support for the scheme. Three quarters of those questioned said that congestion charging has worked. Only 3% of those questioned felt congestion charging has not worked. The remainder did not know whether it had yet worked. However, it is acknowledged that the true impact of the congestion charging scheme on businesses will not be known, with any degree of accuracy, for at least a year.
LONDON LORRY CONTROL SCHEME REVIEW The Mayor's Transport Strategy proposes to review the London Lorry Control Scheme (LLCS), which is commonly known as the London Lorry Ban. The former Greater London Council (GLC) introduced restrictions to stop unnecessary heavy lorry movements through London, to protect the amenity of its residents. LLCS came into force in 1986. It is now administered by the Association of London Government on behalf of the London boroughs. LLCS applies to lorries over 18 tonnes gross vehicle weight. The scheme includes all roads in London except a defined network, called "exempt network" which is designed to prevent through movements. The controlled periods for lorry movements are as follows: (i) between 21.00 and 07.00 on Mondays to Fridays inclusive; (ii) between 00.00 and 07.00 and between 13.00 and 23.59 on Saturdays; (iii) at any time on Sundays. Any goods vehicle over 18 tonnes gross vehicle weight that wishes to use a road that is part of the London Lorry Ban during the controlled hours must be covered by an exemption permit. These permits are available to vehicles that can demonstrate a need to use restricted streets at controlled times.
Developments in urban distribution in London
305
The scheme and its operation has received criticism from the distribution industry from the beginning. Although the operational aspects of LLCS have changed, following several reviews over the years, to make it more "user friendly", the industry believes that it is still too bureaucratic and that with changes to vehicle design in the last decade, it should be possible to do away with the scheme. The Mayor's Transport Strategy proposals respond to these concerns, but bearing in mind the value of the ban to residential communities of London. The review, which was commissioned by LSDP, with the agreement of the members of the Road Freight Working Group which includes industry and local authority representatives, is now underway. The review is being carried out in two stages: (i) (ii)
Stage One will consider the Exempt Network to assess its validity and recommend changes if appropriate; Stage Two will be a more fundamental review and look at the Ban in the light of the new measures and proposals, such as the Congestion Charge scheme, the Low Emission Zones, and the move towards more flexible delivery hours.
The recommendations could have a far-reaching impact on the way lorries operate in London.
Low EMISSION ZONE (LEZ) PROPOSALS Legislation introduced in recent years in the UK is helping to ensure that air quality is monitored at a local level and, if necessary, measures put in place by local authorities to improve it (ALG/GLA, 2002): "Under part IV of the 1995 Environment Act, local authorities have a duty to periodically undertake a review and assessment of air quality in their area. The Air Quality Regulations 2000 prescribe air quality Objectives and the dates for meeting them. For each Objective, local authorities have to consider present and likely future air quality, and assess whether the Objectives are likely to be achieved in time. Should the review and assessment process indicate that one or more of the air quality objectives will not be met by the prescribed date in either whole or part of the authority, and where people are exposed for the relevant averaging time, local authorities are required (after due consultation) to designate the area of pollution exceedance as an Air Quality Management Area (AQMA). They then have a duty to consult on and prepare an action plan laying out plans to work in pursuit of the Objectives." One important measure that could help improve air quality in urban areas is the implementation of a Low Emission Zone (LEZ). The aim of a LEZ is to improve air quality by excluding older, high-polluting vehicles from specified urban areas and encouraging the faster take up of more modern, cleaner vehicles. A range of vehicles could be subject to a LEZ including light and heavy goods vehicles, buses, coaches, taxis and cars. There are no LEZs currently in operation in the UK. However a feasibility study for the introduction of a LEZ in London has recently been carried out (AEA Technology Environment, 2003). This was jointly commissioned by the Mayor of London, the Association of London
306
Logistics systems for sustainable cities
Government and two central government departments to investigate ways of improving the quality of air in London. A significant proportion of the UK goods vehicle fleet enters London during the course of a year, and would therefore need to comply with an LEZ scheme. The expected air quality benefits resulting from a London LEZ scheme would be likely to rise in accordance with the geographical area over which the scheme was applied. The results of this feasibility study are currently being considered by the organisations that commissioned it, and they will decide what action to take with respect to a LEZ in London.
CONCLUSIONS The distribution and logistics industry in the UK is undergoing a great deal of change. There is now considerable interest in the application of principles of sustainability to freight movement and distribution. Companies are also becoming more actively involved in programmes concerned with corporate social responsibility. The EU regulations on energy and air quality, Government interest in operational efficiency, cleaner fuels and fleet modernisation programmes, and operators desire to maximise efficiency through performance measurement and improved utilisation are all contributing to this change. The increasing complexity of operations in large and growing cities such as London place considerable pressures on distribution companies. The new governance in London, with the election of its first Mayor in 2000, has brought a new vision to transport in the Capital. The implementation of the Central London Congestion Charge and increased investment in buses are two prominent examples. Changes in distribution and deliveries have not been as prominent so far. However, the creation of a 'Logistics Platform', the London Sustainable Distribution Partnership (LSDP), with a clear focus on delivery, is a major step towards tackling the short and long term problems that affect the movement and delivery of goods in London. LSDP, whose members are drawn from the distribution industry, local authority, business and amenity groups is a new mechanism to reach consensus on difficult cross cutting issues and improve the sustainability of freight. In the short time it has been in existence, LSDP has already become a focal point for enquiries and proposals on freight and is actively encouraging some initiatives through funding and political endorsement. Through the work of the LSDP in the past year, it has become clear that the lack of detailed information on freight movements in London could become an obstacle to policy formulation in this field. Although commercial information exists, this is not readily available to policy makers. The lack of information has been highlighted as a problem area and it will be tackled, although this could take some years. Similarly there is no freight model for London or the region and TfL and Government are together considering the development of such a model. In the short term, information from Congestion Charging cameras will be useful for central London, at least on a micro scale. London is in the fortunate position of having a new governance with the task of producing strategies for transport, development and environment, which will guide the growth in the Capital for the next 15 years. Freight has often been seen as the 'Cinderella' of transport strategies in the past but this has changed with the Mayor's Transport Strategy which has given
Developments in urban distribution in London
307
it the recognition it deserves. The importance of an effective distribution industry for London has never been in doubt but until recently no structure has existed to take an overview of its operation and problems. There is growing confidence that the challenges of today and the future will be met through the partnership of industry, London boroughs and TfL, under the umbrella of the LSDP.
REFERENCES AEA Technology Environment, Kings College London, Transport Research Laboratory, Transport and Travel Research, University of Westminster and Acona (2003). The London Low Emission Zone Feasibility Study: A Summary of the Phase 2 Report to the London Low Emission Zone Steering Group, http://www.london-lez.org/ Association of Local Government and the Greater London Authority (ALG/GLA) (2002). London Low Emission Zone Feasibility Study: Phase I Report of the Steering Group, http://www.london-lez.org/ Department of Environment, Transport and the Regions (DETR) (2000). Transport 2010: The Ten Year Plan, DETR. Freight Transport Association (February 2003). London Congestion Charging: A guide for vehicle operators, FTA. Greater London Authority (2001). The Mayor's Transport Strategy, GLA. Greater London Authority (June 2002). Draft London Plan, GLA. Strategic Rail Authority (SRA) (2001). Freight Strategy, SRA. Transport for London (June 2003). Central London Congestion Charging Scheme: Three Months On, TfL. Transport for London (TfL) (2001). Transport Statistics for London 2001, TfL.
This page is intentionally left blank
22
AN EXPERIMENTAL COOPERATIVE PARCEL PICK-UP SYSTEM USING THE INTERNET IN THE CENTRAL BUSINESS DISTRICT IN TOKYO
Toshinori Nemoto, Hitotsubashi University, Tokyo, Japan
ABSTRACT Inefficiency in urban freight transport is partly resulted from a number of uncontrolled tracks of different carriers entering the business districts, in order to quickly respond the shippers' pick-up requests. To improve this situation, a co-operative pick-up system by the internet was experimented in Tokyo downtown area, in an effort of public-private partnership involving stakeholders concerned, hi the experiment, we tried to standardize the data and messages and proposed an internet system to match transportation demand (cargoes) and supply (trucks) automatically, expecting more convenient ordering for the shippers, less frequent pick-ups for the carriers, and less congestion for the society. It was found that a small number of shippers used the system since the participating shippers were doubtful of its convenience, while the system did not cause any technical problems. The economic analysis, however, suggests that we could introduce favourable conditions to increase the participation of shippers and utilisation of the system.
INTRODUCTION In Japan, commercial transactions through the internet between individuals and companies, and among companies, have intensified since cheap high-capacity communication began to be available. These transactions have directly impacted the form of distribution by increasing the frequency of small-lot deliveries from upstream business to downstream business (or from producer to final consumer, in the extreme case). As a result, the number of cases of parcel delivery companies responsible for moving the goods has been increasing.
310 Logistics systems for sustainable cities Although parcel delivery companies are efficient in practising full-load transport of large trucks between cities, the delivery and pick-up of cargo within the city is relatively inefficient, thus causing traffic congestion. In particular, a number of small trucks park on the roadside during delivery in the commercial business districts because several building establishments do not have an underground parking space wherein delivery trucks can load and unload their goods, or because the amount of shipment may be so small that parking off the road is a troublesome thing to do. The picking-up of parcels is also inefficient. In the central business district where a large amount of door-to-door transport is generated, there are cases when small trucks used for picking-up goods wait on the roadside so that they can immediately respond to unexpected requests from shippers. To improve the above situation, the Ministry of Land, Infrastructure and Transport conducted an experiment in Otemachi, Tokyo in 2002. In this experiment, transport requests by shippers are made easily through the use of internet (shipper's merit), and an appropriate logistics service provider collects the bundled transport demand for each building, thus increasing transport efficiency (carrier's merit) and reducing roadside parking and truck traffic in terms of vehicle-kilometres resulting in less congestion and environmental burdens (social merit). The consolidation of pick-up activities, which is one of the challenges in the experiment, is more difficult than that of delivery activities. Consequently there have been few successful cooperative pick-up practices while consolidated delivery systems are working in several cities including Fukuoka (Nemoto, 1997), Monaco (Taniguchi and Nemoto, 2001) and Nurnberg (Planung Transport Verkehr AG, 2002). Another challenge was the establishment of a planning committee to design and implement the experiment, involving shippers and logistics service providers in the Otemachi district, as a public-private partnership (PPP). PPP has been introduced to ensure stakeholder participation in freight transport planning, for example in the UK cities (Department for Transport, 2003). In our view, an experiment would be an effective step for successful PPP by sharing common understanding on the impacts of freight transport policies and by eliminating stakeholders' unfounded concerns. The objective of this paper is to clarify the following: a) the principles of the experiment based on the review of E-logistics, b) the detailed system design carried out by the planning committee for its implementation, and c) the results of the experiment and favourable conditions to promote co-operative parcel pick-up system using the internet. The paper will then try to present conclusions.
E-LOGISTICS AND EXPERIMENTAL DESIGN The E-logistics model The mechanism that matches the shipper's demand and the logistics service provider's supply through the internet has been in existence and is termed as "E-logistics". In this section, the existing structure of E-logistics was first examined to understand important concerns that shippers and logistics service providers might have (Figure 1). Type A is the bulletin board type wherein shippers and carriers (or logistics service providers) freely input cargo information and empty truck information. After checking the bulletin board,
An experimental cooperative parcel pick-up system using the internet
311
interested shippers and carriers negotiate separately by telephone the various terms of the contract, including freight charges. This type of scheme allows some website administrators to charge commissions from the shippers and carriers while others profit from other means. The former involves charging of a fixed commission after both parties have agreed on the service contract, for example 5% for freight liquidation and insurance. The latter enables websites to earn income through banner advertising and from utilising the site for manpower staffing and selling of office equipment. Type B is the auction type wherein empty truck information registered by the carriers is offered by bidding to several shippers. The highest bidder will then get the contract for the logistics service offered. This type of scheme is difficult to implement in practice because it needs to involve a certain number of potential buyers requiring the same conditions to join the auction before automatic matching is attained. Stated conditions not only include the trip origin and destination but also the time for the transport service and the type of vehicle.
Figure 1 Business models for E-logistics (Krishnamurthy, 2003 and Daiwa Research, 2001)
312 Logistics systems for sustainable cities Type C is the reverse-auction type wherein the cargo information and the desired price for the service are registered by the shippers. The contract is then awarded to the carrier who first accepted the price and signifies their intention to perform the service. This concept is more difficult to implement in practice than Type B because it requires a certain number of potential sellers out of relatively small number of willing shippers in a depressed economy at present. Type D is the exchange type wherein sellers and buyers, who simultaneously share information and conditions, search for their appropriate partners. Automatic matching is first utilised to reduce the number of candidates who satisfy the basic conditions. Then these conditions are subsequently adjusted manually. A good example, which can be regarded as successful, is the site by NTE Inc. (formerly the National Transportation Exchange), a non-asset third party logistics provider. The company is responsible for the determination of freight charges and other payment matters. As indicated by Hayashi (2002) that even in the US, the slowdown in the progress of IT caused the recent economic recession and therefore IT's serviceability is being reassessed in the field of logistics. Even the excitement brought about by the logistics e-market place (E-logistics), such as trade exchange using the internet and truck and cargo matching systems, also experienced decline and are likewise being reassessed. Truck and cargo matching system by telephone and by internet From the point of view of business, there are only a few successful E-logistics websites in Japan. Some are successful in getting a number of contracts, however, including those who only use the internet to arrange payment of freight charges and facilitate documents for insurance. The case is, in fact, an example of Type A where actual matching is performed by telephone. Misui (2002) pointed out that there were almost 50 entities that employed the internet truck and cargo matching system from its inception by Japan Digicom, Inc. at the end of 1996. These entities not only include logistics service providers but also other companies, such as shippers, information system companies, trading companies and business ventures. Later, however, several companies, including the Japan Digicom, Inc., announced withdrawal from business or shutdown of their truck and cargo matching systems. In order to identify the reasons why E-logistics is quite unpopular in Japan, Japan's transport industry, in general, and the truck and cargo matching system, in particular, will be investigated. First it must be recognised that majority of the logistics service providers are also the shipper or the client (in over 70% of freight contracts). This means that there exist two or three hierarchical structures in Japan, the so-called principal contractor, the subcontractor and the sub-subcontractor. From the viewpoint of the shipper, they experience difficulty in requesting service from a logistics service provider in the area where the cargo is to be delivered. It could be convenient if they collectively commission all cargo to a credible logistics service provider associated with them. This concept is the so-called "one-stop services" for shippers or "third party logistics" for logistics companies who subcontract the services.
An experimental cooperative parcel pick-up system using the internet
313
The relationship between the principal contractor and the subcontractor is, in a sense, economically justifiable. Subcontractors could be set-up even if the freight charges are somewhat cheap, as long as a fixed amount of cargo is maintained without a large degree of marketing. Even if the principal contractors get the brokerage fee without doing the actual transportation, they bear the burden of risk of being the guarantor not only for the operating expenses but also for the quality of work of the subcontractor. The relationship between the principal contractor and the subcontractor is maintained by keeping this delicate balance. The truck and cargo matching system using the telephone in Japan thus evolved to become a mechanism facilitating mutual requests of transport services among logistics service providers. It was assumed that the internet truck and cargo matching system would replace the conventional system of using the telephone. With the internet, the principal contractor seemed to have more chances to meet cheaper subcontractors, while the subcontractor seemed to have more chances to meet principal contractors who are willing to pay higher freight charges. Unfortunately, however, the internet matching systems have not become popular because they lack a mechanism to mitigate against the risks involved, which relied mainly on conventional human relationships in the past. The principal contractor is afraid of failure in the deliver service when the cargo is entrusted to an unknown logistics service provider with cheap freight charges. The subcontractor is afraid of failure with respect to receiving the right amount of payment contracted for freight charges. It is difficult to reduce the risks involved in truck and cargo matching systems. Risks must be mitigated against as much as possible. If these risks were reduced systematically, the internet system would become attractive. To cite an example, a risk assessment system could be developed, as similar efforts are being made with increasing security concerns. Another practical way several websites adopt is a membership system that only allows participants who could fulfil predetermined qualifications. Direct shipper participation in matching We have another assumption that the internet matching system involving the shippers directly is more effective, which is not common yet in Japan. If shippers can directly join the bidding process, the more desirable solution could be obtained by increasing the flexibility of finding an appropriate shipper-carrier combination making efficient configuration of transport, consolidation (de-consolidation) and storage. Though it is convenient for the shippers to request service from the same carrier regardless of the cargo's destination, the logistics chain might be sub-optimised because of the constraints concerning transport, consolidation and storage they or their affiliated subcontractors have. Risk management is more important for the shipper-participating in an internet matching system than the inter-carrier matching system, because shippers (carriers) have not conducted business with many carriers (shippers) and they do not know each other well. One way to reduce risk is to identify the client as credible shippers in a particular district, and at the same time to carefully choose credible carriers that have been doing business there, thus creating a virtual business community. Though the limited membership in a particular area can become a preventing factor for the efficient implementation given the limited extent of
314
Logistics systems for sustainable cities
matching, this disadvantage might be outweighed by the advantage of direct shipper participation. In the experiment, a virtual business community is intended to be formed. It is notable that the experimental system has a weak point because the target of matching is not freight transport characterised as 'inter-urban Full Truck Load (FTL)' which is common in the existing market, but 'urban Less-than-Truck Load (LTL)'. When matching trucks and cargo with inter-urban FTL, it is easier to get a better business opportunity for carriers. Depending on the transport demand, carriers may offer lower freight rates to prevent their trucks from going returning without any backhaul. The internet system enables them to check their partners in a short period of time. However, when matching truck and cargo in urban LTL, they have less incentive to offer lower/higher freight rates manually in real time, even though in the long term they are making every effort to raise the volume of parcels in order to enjoy a higher economy of scale in the production of urban LTL transport. In order to make feasible Elogistics for urban LTL services, transaction costs should be reduced dramatically, for example, by automatic bidding, matching, contracting and dispatching.
Principles of the experiment The experiment was conducted in front of the JR Tokyo station in Otemachi, a commercial business district where rents are very high and only companies of good standing are occupants. Ninety-nine buildings are erected within the 110-hectare area. A major real estate firm has control of the majority of the buildings inside the district. The tenant companies, as members of the District Redevelopment Project Council for Otemachi, Marunochi, and Yurakucho (www.lares.dti.ne.jp/~tcc/index.html), developed a joint project for the area. The shippers in this area have good credibility. Carriers doing parcel delivery business in the same district were invited to collaborate on the experiment. Carriers that agreed to co-operate were registered. The matching process, through the internet, was planned with the objective of seeking the most optimal combination of providing cheap services from the viewpoint of the shipper, allowing efficient pick-up from the viewpoint of the carrier, and reducing environmental externalities for the society.
DETAILED DESIGN OF THE COOPERATIVE PARCEL PICK-UP SYSTEM The detailed system design was entrusted to a planning committee composed of representatives from the shippers, carriers and governmental officials, chaired by Professor Nemoto of Hitotsubashi University. Several problems identified during the process of discussions caused the first draft of the plan to undergo considerable revisions (Table 1).
Matching Bundling Invoice Communication
Table 1 Original plan and implemented system Original plan Implemented system Automatic matching by price, truck Transmitting pick-up requests to location and other conditions fixed pre-contracted carriers Bundling of different carriers' Bundling the same carriers' orders orders with less frequent pick-up Standardised invoice Individual invoice attached additionally Internet, EDI and fax Internet
An experimental cooperative parcel pick-up system using the internet 315 One of the major problems identified was the issue of automatic matching. As mentioned earlier, the objective of the matching process through the internet was to seek the most desirable combination of providing cheap services, allowing efficient pick-up, and reducing environmental externalities. Then the inclusion of freight charges as a factor in the matching process was proposed. This, however, received strong opposition from the carriers because the plan may lead to increased competition in the form of price cutting, which forced the committee to revise the original plan. Although the common household is charged a flat rate for the parcel delivery service, the freight rate of business shipment is negotiated according to the number of transactions in a month and other requirements. Evidently the shippers have the advantage of using the website in finding carriers that offer cheaper freight rates. However, the carriers already doing business in the area insisted that the use of the website does not warrant better business performance (e.g. winning new shippers). As a result, it was decided that the setting of freight charges between shippers and carriers would not be dependent on the experimental system. Even if each shipper requests different carriers for the pick-up service, it is possible that a particular carrier could perform the actual collection work jointly in a particular area. Unfortunately, however, the carriers opposed the method of joint collection. They stressed that the collection must be from the same carrier because the collection provides a good opportunity for their sales drivers to find additional business opportunities. As a compromise, it was resolved that pick-up efforts shall be made by the same carriers and shall follow a system in which for one day, a total of three pick-up times (11:30, 14:30 and 18:00) be designated, not a system in which carriers enjoy the freedom to collect any time they wish. There was also the issue of standardising the invoice forms. The standardised invoice was not enough to reflect the differentiated services of the carriers so that separate invoices of some carriers had to be attached to the standardised invoice forms. Examples of differentiated services offered by the carriers include a service that enable shippers to validate the computerised signature of the accepting person and a record of guarantee of up to 300,000 yen, amongst other features. Furthermore, although a standard system for internet communication was preferred, it was decided that ordinary EDI and FAX were also to be used because of the large investment involved and short duration of the experiment.
316
Logistics systems for sustainable cities
Figure 2 Implemented co-operative parcel pick-up system RESULTS OF THE EXPERIMENT AND ITS EVALUATION Implementation of the experiment The experiment was conducted on weekdays from 28 January to 15 March 2002. The system was installed for 70 shippers (from 14 buildings) who signified their intention to participate in the experiment. This number consists of 0.2% of all the companies that perform business in the district. The participating carriers include the top five in the parcel delivery business, Sagawa Express, Seino Transportation, Nippon Express, Fukuyama Transporting, and Yamato Transport. However, the number of shippers that used the experimental system to request pick-ups was very small. Only 12 shippers out of 70 used the system. In addition, the total number of parcels did not exceed 256 pieces, or 8 pieces per day on the average. Thus, the collection efforts of carriers three times a day did not result to an efficient pick-up system.
Figure 3 Number of collected parcels by time period
An experimental cooperative parcel pick-up system using the internet
317
Evaluation Questionnaire surveys were administered to the participating shippers and interviews were conducted for the carriers after the experiment. The internet system was basically well operated by the shippers and they indicated that they did not encounter any technical problems. At the same time, however, there was a lot of negative replies to the question, "Is the system more convenient or not?" The reply "It got complicated", received a higher percentage than "It became convenient" (Table 2). It was believed that the use of the internet offers little benefits in cases when the amount of parcels per request is small. Table 2 Evaluation from shippers: Convenience of the pick-up system It became convenient 3 (15%) No change 4 (20%) It got complicated 6 (30%) I do not know 7 (35%) Total 20 (100%) It can be said that integrating the pick-up process to three times a day resulted in a reduction in the service quality. In the present practice of pick-up by telephone request, a quick response can be expected usually. Even if the carrier cannot respond at once, the client is informed of the pick-up time by an operator orally, thereby having a sense of security, which suggests the importance of adding a call centre function for improving the internet system in the future. Another inconvenience indicated by the 84% of shippers was the inability to designate the time window of the delivery to the consignee. This service is commonly provided without additional charge by major parcel delivery companies. The co-operative system could provide this service when it can further access the management system of the participating companies in order to judge the possibility of the time window requested. In this way, use of the internet system is thought to have few economic benefits, and thus have little impact on increasing the amount of parcels being handled. Although some problems were experienced by the carriers, including, the list of pick-up addresses of shipments did not fit the present system for scheduling vehicles, the standard invoice used was not enough to cover the various services offered by some of the carriers; there was no complaint about the reliability and operability of the system itself. Technically, a mechanism can be developed to add the original services of each carrier in the standardised system. In this experiment, the amount of handled parcels for pick-up was very small, and there were no significant curtailment effects such as the reduction of pick-up vehicles. However, there was an indication from the carriers that benefits can be obtained if the range of geographical coverage and the number of shippers was increased. Furthermore, they have also pointed out that this will be effective if one staff can stay in the building and be in-charge of the collection system at all times, playing a complementary role to the internet system. Pick-up by telephone has been the current practice. However, because of the inefficient collection of cargo, in which carriers sometimes go back and forth to the same building to
318 Logistics systems for sustainable cities collect cargo again after just a few minutes, they seem to share awareness that there is certainly a need to improve the present pick-up system. Co-operative pick-up system with externalities internalised The internet matching system proved to be technically feasible but economically unattractive. This experiment, however, was conducted given the existing institutional and regulatory arrangement, which can be relaxed when examining the economic nature of the system theoretically. We would like investigate the hypothesis that the internet matching system would bring economic benefits to the shippers, the carriers and the society under different arrangements, and to demonstrate it in another experiment in the future. For this purpose, an economic model was developed (Figure 4), where social costs are introduced. In order for all stakeholders to obtain economic benefits to some extent, one of the necessary conditions that the co-operative parcel pick-up system must fulfil is to reduce the total social costs. Here the social costs include the order costs of shippers, pick-up costs of carriers and traffic congestion costs. Also we assumed that pick-up costs are passed on to shippers by way of freight charges, and that freight charges and order costs comprise the cost for shippers. Despite the fact that the current collection activity causes traffic congestion, carriers have not been burdened with the traffic congestion cost, which others have come to bear, and shippers are not even held responsible for this problem. Certainly, the carriers alone should not be blamed. Among the causes of congestion is the lack of off-street parking spaces for the loading and un-loading of goods. Carriers have no alternative but to park on the street, whereas the police cannot strictly enforce the control measures. At any rate, there is no point or logic in comparing the cost of specific stakeholders with and without the co-operative pick-up system if we ignore the existence of external diseconomies. On the contrary, if external diseconomies are to be internalised, it is possible to make a correct comparison of the impact of introducing the co-operative pick-up system. Here, a way of internalising external diseconomies is, for example, to strictly restrain on-street parking and to permit trucks to park off-street with parking fees. Fortunately, the government of Tokyo has compelled new buildings (more than 2,000 square meters) to provide off-street parking spaces for loading goods since October 2002. Taking into account of the advancement of automatic vehicle identification technology as well, the environment for controlling illegal parking has gradually been improved. Traffic congestion cost will decrease and the cost of carriers will increase if vehicles are forced to use the off-street parking spaces with parking fees. Carriers will then be obligated to increase the amount of cargo to be picked-up per stop. It is under these conditions that cooperative pick-up system will have a significant impact. Social costs can be reduced as external diseconomies are internalised. Although carriers have to pay new parking fees, which can be the cause of increased prices for services, it is eventually expected that parking fees can be saved and prices are reduced since the introduction of a co-operative pick-up system with capability to consider truck location will result to efficient vehicle load usage. Even though there is no assurance that prices will
An experimental cooperative parcel pick-up system using the internet 319
Figure 4 Social costs of pick-up with/without Co-operative Pick-up System (CPS) decrease, it is possible to reduce the costs for shippers, composed of ordering cost and service price. The co-operative system will not become popular if these conditions are not realised. Other favourable conditions Two additional conditions deserve examination. The first condition is the impact of the introduction of co-operative pick-up systems inside the building. These systems have been introduced in some buildings in Otemachi such as in Mitsui Trading and Shin-Otemachi buildings to reinforce security by performing joint receiving and delivery inside the buildings. If this becomes successful and common, the shippers would be ready to accept the area-wide co-operative system as the extension of in-building systems. The second condition relates to the co-ordination of co-operative pick-up with delivery. As clarified by our surveys, some of pick-ups are requested in the morning, when deliveries are usually carried out. Good collaboration between pick-up and delivery will result in even more efficiency. Although the problem is much more complicated, there is a need to investigate this co-ordination. By combining these conditions, it can be shown that "reduction in social costs", "reduction in carrier costs", and "reduction in shipper costs" are highly probable impacts of the implementation of a co-operative parcel pick-up system using the internet. Although it was discussed that it is necessary to reduce the cost of shippers to encourage and stimulate demand, it is also crucial that the carriers should have benefits. The perceived benefits are less for the carriers. In our view, the carriers would be compensated with increased amount of freight cargo through the reduction of freight charges. This is because profits would increase in proportion to the amount of freight cargo supposing reasonable profit per parcel.
320
Logistics systems for sustainable cities
CONCLUSIONS An experiment on co-operative parcel pick-up system was conducted as an effort of PPP in freight transport planning. It did not caused any technical problems while auto-matching was not fully implemented. Only a small number of shippers used the system since the participating shippers were doubtful of its convenience, and there were no reduction in the freight charges. The carriers, on the other hand, indicated that the system could be of use if the amount of freight cargo for pick-up increased in future. These results do not imply that the co-operative parcel pick-up system in the urban area should be viewed negatively. It was found that the limited conditions during the experiment resulted in the low utilisation of the system, and thus, it is necessary to examine favourable conditions to increase the participation of companies and utilisation of the system. The economic model of social costs of co-operative parcel pick-up system suggests that reducing the total social costs would be achievable, and that all stakeholders including the shippers, the carriers and society could obtain economic benefits to some extent with the introduction of co-operative parcel pick-up system if the external diseconomies were internalised by enforcing off-street parking, for example. If we could increase our common understanding of the economic nature of the system in a future PPP attempt, all the stakeholders would be willing to join the internet co-operative system.
REFERENCES Daiwa Research (2001). Development of the Truck and Cargo Matching System. Keiei Joho Search, Summer 2001, 32 (in Japanese). Department for Transport (2003). A guide on how to set up and run Freight Quality Partnerships, Transport Energy Good Practice Guide 335. http://www.freight.dft.gov.uk/gpg355/pdf/gpg335-final.pdf Hayashi, K. (2002). Changes of IT and Logistics in the US after the Recession. In: IT's Impacts on Logistics, The Japan Research Center for Transport Policy (in Japanese). Krishnamurthy, S. (2003). E-Commerce Management- South-Western, Thomson Learning, Mason, Ohio. Ministry of Land, Infrastructure and Transport (2002). Report on Advancement in Road Transport Using IT (in Japanese). Misui, Y. (2002). Truck and Cargo Matching System Using Information Communication Technology: Current Conditions and Future Issues, Proceedings of the 17th Annual Conference of the Japan Association of Social Informatics (in Japanese). Nemoto, T., J. Visser and R. Yoshimoto (2001). Impacts of Information and communication Technology on Urban Logistics System. In: OECD Web conference proceedings; http://www.oecd.org/oecd/pages/home/displaygeneral/0,3380,EN-document-25nodirectorate-no-20-19078-25,FF.html. Nemoto, T. (1997). Area-wide inter-carrier consolidation of freight in urban areas, Transport Logistics, Vol.1, No.2. Planung Transport Verkehr AG (2002). Innovative Distribution with Intermodal Freight Operation in Metropolitan Areas, Funded by EC. http://europa.eu.int/comm/transport/extra/final_reports/integrated/IDIOMA.pdf Taniguchi, E. and T. Nemoto (2001). City Logistics—Efficient and Environment Friendly Urban Freight Transport Planning, Morikita, Tokyo (in Japanese).
23
NEW IDEAS FOR THE CITY-LOGISTICS PROJECT IN KASSEL Uwe Kohler, University of Kassel, Germany
ABSTRACT The city logistics project Kassel, which was started in 1993, was successful for a number of years. But in the recently, little interest has been shown in the project. Therefore, an analysis has been carried out to find out the reasons for this decline in interest and to determine the factors necessary for a successful city logistics concept. A new city logistics project is to be planned, taking into account the analysis carried out on other German city logistics projects. In contrast to the earlier concept, the new one is to be based on a specialised form of organisation which will embrace more closely the interests of all those involved. Furthermore, the concept is to be extended and improved by installing modern route planning software. A freight traffic centre, currently under construction and an expansion of the city logistics services will be included in the new project.
CITY LOGISTICS IN KASSEL FROM THE BEGINNING TO NOW The freight transport initiative in Kassel, frequently termed, the "city logistics project Kassel", was among the first generation of this kind of project to be carried out in Germany. The aim of these early efforts was deliberately to reduce the constantly rising traffic volume in cities, especially industrial and individual traffic. In 1993, an association of forwarding agencies and the Chamber of Commerce and Industry established a committee to bring together interested forwarding agencies who supplied the city centre of Kassel. The preparatory and introductory phase of the city logistics project Kassel was accompanied scientifically by the University of Kassel's Department of Traffic and Transport Planning. This included preliminary research work, such as surveying retail traders, surveys of forwarding agencies
322 Logistics systems for sustainable cities and cordon counting before and after starting the project. In collating these research results, the concept of the freight transport initiative "city logistics in Kassel" was established. The beginning of the city logistics project in Kassel
Main features and effects From August 1994, ten forwarding agencies participated in the co-operative system. Large agencies with a nationwide branch network collaborated, as well as smaller agencies with less than 20 lorries (StrauB 1997:68). A neutral carrier was employed to collect and distribute the bundled goods in the centre of Kassel, a defined area which included the pedestrian zone and adjacent roads. Early in the morning, the neutral carrier collects the goods bound for the centre of Kassel. These goods are delivered during the night to the forwarding agencies. At a terminal near the city centre, the goods are commissioned. In the morning, the goods are grouped according to receivers' addresses and road routes, are loaded onto two to three specially labelled trucks. Usually, a truck drives two routes per day, the number of vehicles and trips depends on consignment volume. Using this procedure, the technical effort needed for preparation, billing and calculation is kept low. In the morning, the forwarding agencies inform the neutral carrier via e-mail about the quantity of goods which have to be collected. The collection, commissioning, bundling and distribution are done manually or by means of receipts which the neutral carrier hands over to the forwarding agencies. Barcode technology has not been used up to now, because the forwarding agencies often use different systems (Klein-Vielhauer 2001:19). The main features of the city logistics project in Kassel since 1997 and today are: (a) (b) (c) (d)
a neutral carrier is employed to collect and distribute the goods it relates to a defined city area of Kassel, not only to the supply of difficult customers additional handling is done in the city-terminal to bundle the goods, it works without a freight traffic centre
Effects on bundling traffic. Before the co-operation began, separate deliveries meant that forwarding agencies covered an average distance of 118 km to the city centre per day, while the neutral carrier need just 69 km per day. This represents a reduction of around 40%. By means of the co-operation and improved route optimisation, the forwarding agencies achieved a traffic reduction of approximately 60% (from 25km down to 10 km). The number of trucks driving into the city per day was thereby reduced from 15 to 2 or 3 trucks with 4 to 6 trips, which means that 11 truck trips per day were saved. Model calculations show that vehicle utilisation was improved by the bundling of the goods by the neutral carrier. It was shown, that the loading volume for deliveries to the city centre of Kassel improved from 40% without city logistics up to 80-90% with city logistics. The vehicle utilisation with respect to weight improved from 25% to 60%.
New ideas for the city-logistics project in Kassel
323
Effects on city traffic. The forwarding agencies who took part in the city logistics project represent 3% of Kassel's total freight traffic (StrauB, 1997:72-82). Therefore, the reduction of Kassel's total traffic volume through the introduction of city logistics was only marginal (Kohler, 2001:6). Economic effects. The original intention of the city logistics project was not only to reduce traffic but also to increase the economic efficiency of deliveries of the centre of Kassel. Whether this was actually achieved can only be estimated, because the conventional cost accounting of the forwarding agencies make no calculations regarding the costs for individual deliveries but only for complete deliveries. The changes in costs for the co-operation of forwarding agencies in the city logistics project of Kassel can be described as follows: before the co-operation was realised, the forwarding agents delivered the goods with their own vehicles. The fixed costs for vehicle maintenance, personnel costs as well as operating costs were calculated, or sub-contractors took over this task. The final payments were based on tariff or lump sums. Since the beginning of the city logistics project, a neutral carrier has taken over the deliveries in the city centre. Due to collection, storage, handling and distribution, additional costs arose which were shared among the forwarding agents. On the other hand, the vehicles and personnel resources which are not involved in carrying out deliveries to the inner city are free to take on other tasks. It is hardly possible to obtain an exact comparison between saved costs and additional costs because of the difficulties caused by identifying the costs on a consignment basis. Possible cost savings are offset by additional handling procedures and the resultant higher transport costs. So it can be taken that the cost advantages and disadvantages of city logistics of Kassel are evenly balanced and that the intended cost reductions were not achieved (Kohler, 2001, p. 6). The economic benefits of the city logistics of Kassel are therefore marginal. So other advantages have to be emphasised, in order to justify the continuation of the project (Straufi, 1997: 84-85). Figure 1 shows the way the city logistics project in Kassel currently works.
324 Logistics systems for sustainable cities
Figure 1 City logistics project in Kassel today
Current state of city logistics in Kassel After operating for eight and a half years, there have been a number of changes compared to the original situation at the beginning of the project. Half a year after the project started, the neutral carrier dropped out Temporarily, two groups were formed from among the 10 co-operating partners. One forwarding agent from each group took over the delivery for the city centre for the remaining members of the group. After quite a time, a new neutral carrier was found, so that from the beginning of May 1996, the original city logistics concept started again. In the course of time, three of the 10 partners left the co-operation because of changes in their area of business or because of an insufficient quantity of goods for deli very to the city centre. Consequently, just 7 forwarding agents and the neutral carrier take part in the city logistics in Kassel today. The nature of the procedures at the technical-physical level have not changed up to now. Modem information or communication techniques (e.g. barcoding, electronic route planning systems and telematic services) are still not used.
New ideas for the city-logistics project in Kassel
325
Up to now, it has not been possible to realise the original plans for expanding the co-operation of the city logistics concept, such as incorporating further forwarding agencies, co-operating with further freight transport initiatives, like the freight traffic centre Kassel and extending the range of services. Overall, the present situation of the city logistics project in Kassel shows a general decline in interest among all participants.
REASONS FOR THE DECLINING INTEREST IN CITY LOGISTICS KASSEL
The situation nationwide The initial euphoria, with which approximately 200 city logistics projects were planned and carried out in Germany in the mid-90s, has meanwhile entered a phase of disillusionment (Klein-Vielhauer 2001:12). Many projects were stopped soon after the initial phase or somewhat later, as it became obvious that the high expectations borne by the projects could not be fulfilled, with regards to the effects on traffic and the environment as well as their economic consequences. Studies have shown that freight traffic comprises only 10-15 % of the entire traffic in the inner city. According to surveys, city logistics projects cover only 2-5 % of the total freight traffic volume (Holser, 2002:1). This percentage is a drop in a bucket, compared to the overall traffic flow. Visible improvements in environmental and traffic development are, however, necessary to obtain (financial) support for the projects from politicians and municipalities. Usually, lacking financial support or running out of it puts an end to city logistics projects. This results from the fact that, at the beginning, cost savings resulting from bundling the deliveries to the city centre were overestimated. Higher expenses for additional handling and transport of goods were, at best, just covered. In general, profits were not made.
The situation in Kassel Even in Kassel, a decline in interest in the city logistics project can be clearly observed. There are various reasons for this. The main reason is the differing interests of the participating forwarding agencies. On the side of the co-operating forwarding agencies, the following problems can be identified: (a) No regular meetings were held by the participating forwarding agents (b) The forwarding agents did not come to mutual agreements and, similarly, the organisational work was not supervised (c) Among the co-operating partners, there was insufficient information exchange about the project (d) The individual forwarding agencies used the co-operation predominantly for delivery orders which were not lucrative. Profitable deliveries are still carried by themselves (e) Up to now, Kassel does not have a freight traffic centre, so logistical services are not available The result of this has been a steady decline in the economic incentives for maintaining co-operation among participating agencies in Kassel. In addition, the necessary pressure to support this cooperation is lacking. Furthermore, the Kassel authorities have not adequately enforced local measures
326
Logistics systems for sustainable cities
such as restricting delivery time (Klein-Vielhauer 2001:23), and the city has no spatial or structural constraints like a historical city centre with narrow streets, places of interest and streams of tourists. A further significant point is the fact that, so far, one of the important participants for city logistics, the local retailers, have shown hardly any willingness to get actively involved in the city logistics project. Even before starting the Kassel project, retailers were critical about handling the delivery of bundled goods to the city centre, about scheduled deliveries and about the large quantities of goods which do not arrive through the day (Straufi, 1997:57). The forwarding agencies offered to deliver outside office hours, but the retailers refused such deliveries for reasons to do with staffing. Up to now, the retailers have not been willing to get the dispatchers of the received goods to authorise only forwarding agencies participating in the co-operation. The retailers normally get their goods delivered free of charge. Therefore, neither in Kassel nor in other cities do they have any great interest in considering the problems of delivery traffic. Increasing obstacles with urban deliveries and the unsatisfactory cost situation for forwarding agencies will have negative effects on the quality of delivery in the long run. The retailers lack a long-term view which is needed if the quality of shopping for end-customers is to be improved. In addition, the city logistics project Kassel has no neutral mediator to deal with disagreements, to give new impetus to further developments and to initiate marketing activities.
DEVELOPMENTS AMONG OTHER GERMAN CITY LOGISTICS PROJECTS Two projects are presented below to portray successful city logistics projects in Germany. The Numbers city logistics model ISOLDE (inner city service with optimised logistical services for the retail trade): May 1996 saw the start of the pilot project of a co-operation of forwarding agencies and participating parcel services, to supply the inner city of Nilrnberg. In association with individuals and institutions from the region's industries, waste disposal companies, retail services, traders and municipalities in the region of Niirnberg, a company named, "Integrated freight transport management in North Bavaria (IGN)" was established. The project work covers five basic components: (a) city line service (b) city waste disposal service (c) city storage service (d) city shopping service and home shopping (e) city marketing service The city line service consists of two elements: the bundling of goods from forwarding agencies and, on the other hand, the packages of parcel services, together with their shared supply. Meanwhile, however, all the participating parcel services have left the city logistics project for financial reasons. Only four forwarding agencies are now participating in the bundling of goods, all of them situated in the central freight traffic centre at the port of Niirnberg. A neutral carrier starts in the morning and following a relatively fixed time schedule, makes trips to the four co-operating forwarding agencies, one after the other. The goods are bundled on the vehicle.
New ideas for the city-logistics project in Kassel
327
Around 9am, they are delivered to an inner city depot or direct to retailers. In addition, smaller, lighter goods are delivered by electrical mobiles from this depot to retailers. Delivery routes are linked to pick-up routes for waste material back to the depot. When the waste is picked up by urban garbage collection trucks. If necessary, shopping goods are carried out additionally by home deliveries from the participating retailers after 5pm (Klein-Vielhauer, 2001:38). Furthermore, shops in the inner city have the possibility of temporarily storing bulk deliveries, extra goods and seasonal articles at short notice. If required, ISOLDE can also take on "value added services", such as price-labelling and commissioning. The city marketing service is to be understood as a shared platform for advertising and publicity for the participants of Niirnberg city logistics. This service includes active lobbying and working with the Press (Klein-Vielhauer, 2001:25).
ResLos®- Resensburs city logistics project In 1996, BMW, the car manufacturer, began a research project on freight traffic in the city of Regensburg. In 1998, the RegLog® concept was implemented. The following seven components are part of the city logistics project: (a) bundling service (b) environmentally compatible distribution of package freight (c) waste disposal service (d) commissioning storage service (e) depot service for end-customers (f) processing service for traders and end-customers (g) home delivery service The bundling service for packaged goods is carried out by seven forwarding agencies participating the co-operation. Additional handling and sorting is done by one of the co-operating forwarding agents who also works as a freight carrier. An external sub-contractor is put in charge of the daily collection and distribution routes. The bundling service is attached to the disposal service for packing material. The originally planned co-operation of forwarding agencies and parcel services has not yet come into being because of the negative attitude of the parcel services (Klein-Vielhauer, 2001:76). The commissioning storage service on offer is directed at retailers in Regensburg as well as at external forwarding agencies who can use open as well as closed storage. At first, a home delivery service was also offered to end-customers, but, due to low demand, this service was soon withdrawn (Klein-Vielhauer, 2001:75).
Success factors of these projects Both city logistics projects, ISOLDE in Ntlrnberg and RegLog® Regensburg, show certain similar characteristics, as listed below:
328
Logistics systems for sustainable
cities
(a) certain pressure caused by restricted traffic conditions, resulting from a historical city centre with narrow streets, numerous historic sights and large numbers of tourists (b) the early, active involvement of all participants, e.g. by participating financially with the operator company (c) scientific support during the planning and implementation of the project (d) the recruiting of a mediator (e) the integration of a freight traffic centre in the city logistics project (f) the willingness of the local authorities to co-operate, e.g. in checking that access restrictions are enforced
REQUIREMENTS FOR SUCCESSFUL CITY LOGISTICS
General factors for success A successful city logistics project requires very careful preparation and planning. In particular, it seems necessary (a) to initiate public debate about traffic problems in the city in order to create the necessary political pressure to do something (b) to identify affected groups and participants at the beginning of planning, and to get groups involved, so as to take account of everyone concerned and to avoid resistance (c) to gain competent promoters and key actors from different groups to support the project (d) for participants, in a round table discussion, to develop a rough target for the city logistics concept The current situation should be surveyed and a strengths/weaknesses profile determined. As no data is normally available, the inner city freight traffic values should be evaluated. By this point at the latest it is recommended to bring in scientific consultation for the project (e) to develop a catalogue of measures to be realised. Cost-benefit analysis should be undertaken in order to arouse participants' interests in city logistics. Based on a neutral-costs calculation of the planned measures, cost savings must be presented on a medium-term basis (f) not only to involve affected participants and groups from official departments and authorities for the realisation of city logistics projects, but also politicians. The chances of success increase when politicians accept responsibility for the project. Apart from making the financial means available, appropriate political resolutions by the city council can also benefit city logistics (g) in the critical planning phase when decisions must be taken about concrete measures, to do public relations work with the means and facilities available. In addition to achieving transparency, the new targets and strategies have to be presented. Information about other, already successful projects will be of great advantage in promoting general acceptance (h) to install a competent mediator during the realisation of the concept who will take care of the conversion and consolidation of the operating business. At the beginning of the project it is advisable to apply, in particular, win-win measures from which all participants can benefit. This makes positive reporting possible and makes easier the realisation of other points of the project (i) to initiate a constant check of the effects and the success of the measures, with regard to the envisaged aim, and the possible effects with other partial aims. Quality control which accompanies the process will make it possible to steer and intervene in the process and continually to adapt and refine the concept (MWMEV, 2000:36) The planning and shaping of the project depends to a high degree on the traffic volume of the freight transport in a regional economy. This, in turn, is largely determined by the geographical situation of
New ideas for the city-logistics project in Kassel
329
the city, the size of its area, population and catchment area, the traffic infrastructure available, the existing mix of businesses and industries and the available innovative technology. Additionally important are so-called "soft success-factors" such as a creative and innovative environment and a basically co-operative attitude among potential participants. These, too, contribute to making a city logistics project successful (Eisele, 2001:4).
Special factors for success Besides these general factors, for a city logistics project, further special factors for success can be identified: (a) the participants involved should be willing to co-operate. It is recommendable to create a firm platform of interests, led by a neutral mediator, to co-ordinate various individual interests. In addition, the mediator takes care of presenting the launch of the project to the public and takes care, too, of its continuing development (b) the choice of a suitable form of organisation for the project is highly important (i.e. a limited company) (c) A freight traffic centre and / or an inner city depot should be available (d) The pollution and environmental compatibility of delivery vehicles plays an increasingly important role (e) The availability of special city logistics services which go beyond the mere bundling of goods. These include: (i) the commissioning storage service for trade customers (ii) a processing service for trade and private customers (iii) a home order and delivery service (iv) the build-up of industry logistics (v) the availability of waste disposal logistics, particularly for packing materials
DEVELOPMENT O F A NEW CITY LOGISTICS MODEL IN KASSEL Based on the experience with the Kassel city logistics model and with models in other German cities, the following elements for an improved city logistics model can be derived. 1. Basically, it is recommended that all project participants should establish a city logistics company which ensures that all information is available for all partners by means of regular meetings. This would help to prevent partners making lucrative deliveries for their own advantage. 2. Furthermore, a neutral mediator should be nominated who can be contacted in case of problems or decisions which have to be taken. The mediator is to be informed about the current company situation and its development That persons knowledge will be useful in further developing the model, in adapting it or in initiating marketing activities. 3. There should be an investigation into the possibility of Kassel City Council supporting the city logistics project by (i) restricting delivery times in the Central Business District (ii) marking special delivery zones just for city logistics vehicles.
330 Logistics systems for sustainable cities (j) For the fast delivery of smaller quantities, flexible motorcycle and bicycle couriers should be employed besides conventional trucks and delivery vans. (k) Further trip reductions can be achieved using software for navigation and route optimisation. The use of a unified barcode scanner system will simplify the tracking and tracing of goods. (1) The planned freight traffic centre in Kassel will allow all the forwarding agents to bundle the goods on the spot, avoiding the need for collection trips. In this connection, the logistics services could be extended. Among these additional services can be included commissioning and labelling work as well as the short-notice provision of storage space and logistical equipment for retailers. So the retail trade can convert expensive storage space in the city area into lucrative sales areas. (m)Not only the delivery of goods should be considered, but also the waste disposal offered by city logistics services. In one process, the deliveries can be carried out simultaneously with the disposal of packing materials. Apart from saving storage space needed for waste, the number of the city centre trips for garbage collection trucks can be reduced. (n) The introduction of industry-specific logistics can be considered, e.g. for the health sector, banks, financial services and public administration. (o) To simplify the handling of consistently recurring transportation and delivery damage, the city logistics company can establish a shared department to process damage-related matters. (p) It is important for city logistics in Kassel to include local retailers. In view of the increasing number of supermarkets locating at the outskirts of the city and in view of the increasing trade volume via internet (e-commerce), the retailers must have an interest in offering more attractive services in the future. What might help here are home order and home delivery services as well as services for end-customers whose time and mobility are limited. The problem of the "last mile", when supplies cannot be delivered to the receiver because they are absent, could be simplified by installing pick-up points, lockers in public places, bus and tram stops, filling stations and multistorey car parks. In addition, city logistics delivery vehicles could be used for a home delivery service in the evening at agreed fixed times. The aim of the efforts which have been described is the improvement of the cost-benefit ratio in favour of the involved partners. The main differences between the current and the new model are as follows: (a) (b) (c) (d) (e) (f)
establishment of a "city logistics company Kassel" nomination of a mediator inclusion of the freight traffic centre in Kassel active participation of retailers and the city authorities city logistics services like waste disposal and home delivery installation of modem techniques, e.g. route planning software
It can be expected that the new city logistics model in Kassel will be realised in the next few years. At the moment, it is not yet possible to quantify the effects of the new model, for example, with respect to the number of vehicle movements. Figure 2 shows the new city logistics concept of Kassel.
New ideas for the city-logistics project in Kassel
331
Figure 2 The new City logistics Project in Kassel
REFERENCES Eisele, A. (2001). Stadtlogistik: Ein Bericht tiber Verzauberung, Ernuchterung und neue Losungen, Aachen, http://www.netzwerk-stadtlogistik.de/download/eisele-2001-12-04.pdf. Holser, T. (2002). City-Logistik - eine Strategie von gestern?, Industrie- und Handelskammer Frankfurt am Main, http://www.ftmikfurt-main.ihk.de/presse/ink-wirtschaftsforum/2002/0208/citylogistik/index.html. Klein-Vielhauer, S. (2001). Neue Konzepte fiir den Wirtschaftsverkehr in Ballungsraumen - Ein Werkstattbericht tiber Bemtihungen in Praxis und Wissenschaft. Institut fiir Technikfolgeabschatzung und Systemanalyse, Forschungs-Zentrum Karlsruhe GmbH, Karlsruhe. Kohler, U. (2001). City logistics in Germany. The Second International Conference on City Logistics, Okinawa, (E. Taniguchi and R.G. Thompson, eds.), Institute for Systems Science Research, Kyoto.
332
Logistics systems for sustainable cities
MWMEV (Hrsg.) (2000). AbscWussdokumentation Modellvorhaben Stadtlogistik NRW. Ministerium fur Wirtschaft und Mittelstand, Energie und Verkehr des Landes NordrheinWestfalen, Diisseldorf. StrauB, S. (1997). City-Logisttk - Ein Instrument zur Verringerung des stadtischen Guterverkehrs. Schriftenreihe Verkehr, Heft 7, Fachgebiet Verkehrssysteme und Verkehrsplanung, Universitat Kassel.
24
A STUDY O N THE SETTING UP O F LORRYDEDICATED ROUTES IN THE BRUSSELS CAPITAL REGION Wanda Debauche, Head of Mobility Unit, Belgian Road Research Centre, Brussels, Belgium
ABSTRACT Traffic regulation is one way to reduce negative effects produce by heavy freight traffic in urban areas. This article expounds the strategy the Brussels Capital Region is setting up to counteract these effects, especially by providing mandatory corridors for heavy freight vehicles and by restricting their access to residential areas. The four components of the strategy are described : • heavy traffic regulations on routes (imposed or recommended) and their inforcement; • the required regulatory and direction signing; • communication and consultation tools to be developped in collaboration with the actors; • the design of roads and structures.
INTRODUCTION At the end of the nineties the Brussels Capital Region, concerned about the quality of life of its citizens, endowed itself with a transportation plan. This strategic plan for transport policy took an approach to both passenger and freight transport. In the field of freight transport, the following strategic themes and actions were defined: 1 • the parking of delivery vehicles'- facilitate deliveries and eliminate double parking, solve the problems caused by the long-term parking of lorries; 2. the journeys of goods vehicles'- integrate lorries better in the overall traffic flow and protect their journeys, reduce the damage to the environment;
334 Logistics systems for sustainable cities 3. logistics'- improve the accessibility and services of the port area, integrate the logistic function of the Brussels Region in European logistic networks, improve the competitiveness of the local transport and logistics companies. In this context it clearly appeared that the administration and the regional political authorities in charge of transport wanted to further channel the journeys of lorries through the urban network, especially by providing corridors that are mandatory for heavy freight transport and give access to the whole region. A study was ordered from the Belgian Road Research Centre to define the appropriate strategy for these routes. This study was to result in a solution that can be operational on the spot both for the carriers and the shippers (i.e. the economic activity of the region) and for the frontage residents (i.e. the quality of life in the city). The objectives of mandatory routes for heavy freight vehicles were consequently defined as: (a) to secure the quality of life in residential areas; (b) to optimise the journeys of goods vehicles on the road network; (c) to preserve road infrastructure that was not designed for heavy vehicles. A steering committee was formed in order to involve the parties concerned in the development of a realistic solution. It includes representatives of: (a) the Infrastructure and Transport Administration of the Brussels Capital Region; (b) the Land Planning Administration of the Brussels Capital Region; (c) the Brussels Institute for Environmental Management; (d) the Port of Brussels; (e) the Brussels Chamber of Commerce and Industry; (f) the Federation of Companies based in Brussels. The next few Sections describe the defined strategy and its four components: 1. Regulations on freight routes (imposed or recommended) and their enforcement. 2. The two components of signing (regulatory and direction signing): types and locations of traffic signs. 3. Communication and consultation tools to be set up with the parties concerned. 4. The design of roads and structures, especially the design of roundabouts, the accessibility of roads, the planning of intersections, the choice of materials, etc. All these aspects were studied in three test areas, chosen according to the following criteria: • first: the presence of a variety of activities generating heavy goods vehicle transport (business centres, restaurants, wholesale and retail trade, etc.). The variety of established activities should allow a wide range of freight vehicles originating from or arriving at these areas; • secondly: a diversified road configuration which is representative of the Region; • finally: a variety of roads of different categories as specified in the Regional Development Plan. The following areas were chosen: a mixed area with dwellings and shops (Pentagone, number 9 on the map below), an area with industrial and logistic activities (port, number 10 on the map), and an office area (Leopold Quarter, number 11 on the map).
A study on the setting up of lorry-dedicated routes in the Brussels Capital Region
Figure 1 Map of the Brussels Capital Region Test areas
335
336
Logistics systems for sustainable cities
REGULATIONS AND THEIR ENFORCEMENT General principle of the regulations This general principle can be represented as follows:
Figure 2 Scheme of the general principle of the access regulations for goods vehicles in the Brussels Capital Region
A study on the setting up of lorry-dedicated routes in the Brussels Capital Region
337
Exceptions have to be defined, especially for: (a) removals; (b) the collection of household waste; (c) home deliveries; (d) postal services; (e) the needs of public and emergency services. This system of exceptions has to be developed and approved by both the Region and the nineteen municipalities of the Region. The issuing of permits and the system of exceptions In 1993, the Brussels Capital Region instituted an environmental permit. This permit is to offer protection against the hazards, nuisance or inconvenience an installation or activity may directly or indirectly cause to the environment or to the health and safety of the population. The final aim is to ensure economic growth without disturbing the social climate of the inhabitants, and to improve the quality of life in the conurbation without shutting the door on any industrial or commercial initiative. An environmental permit is mandatory for the running of so-called "listed" installations. Any device, machine, installation or activity which could cause a hazard, nuisance or inconvenience to the population or the environment and appears on the record of "listed" installations must have a permit, whether the installation be operated by a private or public company, a private individual, a non-profit organisation, etc. All depends on the size of the installation and the nuisance it causes in terms of environment, health and mobility. Applications for permits are examined on the basis of the information provided by the applicant (environment, mobility, etc.), and the permit is delivered either by the mayor and aldermen of the municipality where the installation is located or by the Brussels Institute for Environmental Management. The authority issuing the environmental permit can impose specific conditions for the running of the business. Following this principle, it has been decided to impose a route on vehicles heading for or leaving a business and causing major disturbance in the form of heavy traffic. The authorisation system was conceived as follows: • for listed activities, after a case-by-case examination of the fleet vehicles and the deliveries (frequency, time-profile, on/of-street parking operations...) a route will be imposed or not. If necessary, permits already issued can be updated (depending on the financial and human resources available for this extra work load). As the most disturbing activities are already subject to an environmental permit, it is suggested to restrict the obligation to apply for an access permit - possibly leading to a mandatory route — to listed activities', •
non-listed activities (safety vehicles, waste collection vehicles, etc.) will receive a permit de facto. The de facto permit system will be developed by mutual agreement between the Region and the nineteen municipalities of the Region (police, departments in charge of town planning and roads), in order to adopt a consistent attitude. For instance, it is suggested to grant a de facto permit in the following cases:
338
Logistics systems for sustainable cities
* removals; * the collection of household waste; * home deliveries; * postal services; * access for public and emergency services. In all other cases than the above-mentioned exceptions, a case-by-case examination seems more appropriate, given the complexity of the existing distribution and supply chains. The examination and delivery procedure described above is represented in the diagram below.
Figure 3 Examination and delivery procedure for access permits and mandatory routes for heavy freight vehicles in the Brussels Capital Region
SIGNING
Direction signing It was suggested to give preference to direction signing operating at different levels. Before entering the Brussels Capital Region: before and on the "Ring" motorway Signing on the highways leading to the "Ring" motorway, to have the heavy goods vehicles take the east or west side of the "Ring". Signing on the "Ring" itself to make sure that the vehicle chooses the right exit, that is, the one that is nearest to the final destination. The traffic signs below will be used. They indicate the direction that vehicles heavier than 3.5 t have to take (left, right, straight on in order to exit immediately) for the stated locations (port, industrial areas, etc.), as well as the sequence number of the most appropriate exit. These signs will be separated from the existing signs, to make them more legible.
A study on the setting up of lorry-dedicated routes in the Brussels Capital Region
339
Figure 4 Examples of direction signing on the ring road The installation of these signs before and on the "Ring" located on the territory of the Flemish and the Walloon regions requires an agreement between the three Regions. In the Brussels Capital Region Direction signing at some distance from the destination, fragmented and indicating the direction to primary areas (like industrial areas), secondary areas (like commercial centres) and tertiary areas (like certain companies). Additional direction signing in the proximity of the destination.
Figure 5 Examples of additional direction signing in the Brussels Capital Region This signing method requires an inventory of all the primary, secondary and tertiary centres of activity, and a clear definition of their official names. The installation of ranging poles indicating that one is in the stated area is also planned. As a matter of fact, drivers are often lost because they do not know that they have reached the area they were looking for. Last but not least, special attention is paid to the way back from these zones to the "Ring".
Regulatory signing In the field of regulatory signing, signs placed at the entrance to residential areas will indicate that vehicles of 19 t and heavier can enter the area only if they hold a permit. The only method that can be used is complementary regulation, in the form of a sign with local validity. This new sign will be suited to the category of users allowed or - in this case - prohibited to travel in the area. As the sign only has local validity, both the entrance and exit of the area have to be indicated.
340 Logistics systems for sustainable cities
Figure 6 Regulatory signing used in the Brussels Capital Region Authorities in charge of enforcement Compliance with the routes that may be imposed for listed activities will be checked in the same way as with the other requirements made in the permit. The official in charge of issuing the permit will also be empowered to record offences. Inspectors of building works will have authority to record violations of environmental permit requirements for building sites. Systematic enforcement will be difficult, given the lack of human resources. As a matter of fact, checks are often made after complaints have been reported.
Liability for offences According to the law, the holder of a permit is expected to observe the conditions under which the permit was issued. If they do not, they will be penalised. They will be reported, but will have the possibility to turn against a carrier who has not taken the imposed route. To protect themselves from liability, we suggest that the permit holder should include any existing obligation to take a set-up route in their transport contracts. In doing so, they will have to make sure that the carrier is informed about this route. Remarks This system of regulatory signing could cause some problems: (a) the set-up routes will use roads belonging to different categories (especially in terms of authorities in charge of their management, maintenance and signing). For instance, there will be regional or local roads, and roads crossing several municipalities. Hence, the nineteen municipalities of the Region and the Region itself will have to dialogue and co-ordinate in order to agree upon: (i) the geographic definition of the areas where lorry traffic is restricted; (ii) the exceptions to the restriction. Failure to co-ordinate would endanger the viability of the principle of protected areas; (b) the areas with limited access are numerous. Their delineation will require a considerable number of signs. This is inconsistent with road legibility and efficient signing. As a matter of fact, many road managers agree that signing should be simplified and that the number of signs should be reduced, rather than introducing new ones. Moreover, the budget will increase accordingly;
A study on the setting up of lorry-dedicated routes in the Brussels Capital Region
341
(c) there must be sufficient enforcement to guarantee that the new legislation is observed. However, the authorities and police are complaining even now about inadequate human resources to enforce and check compliance with the existing legislation. Consequently, the Administration has been recommended to: (d) choose a policy commensurate with the resources that can be dedicated to enforcement and penalisation. It will be necessary to work in stages. To begin with, we recommended concentrating on direction signing rather than regulatory signing; (e) endow the authorities in charge of enforcement - of both the requirements in the environmental permit and the parking regulations for vehicles in general - with sufficient technical and human resources; (f) work closely together with the nineteen municipalities of the Region, at both levels of infrastructure managers and authorities (enforcement and policing).
COMMUNICATION AND CONSULTATION TOOLS The success of the setting up of mandatory routes for heavy freight vehicles depends mainly on two factors: first, the dialogue between the players concerned by the various measures, and, secondly, the means of communication used to inform the public about the new rules and measures.
Consultation The study recommends organising consultation at different levels: • continuous general consultation at the regional level on the problems of urban freight transport as a whole1. The aim is to provoke an exchange of ideas and to confront the points of view of the infrastructure managers, the shippers (traders, industries, etc.) and the professional carriers, represented by their federations. This would allow apprehending the problem independently of mere local considerations or interests. The different levels of power (federal, regional, municipal) involved in infrastructure management and transport regulations should be gathered. Such an approach would show the interest of the regional authorities in freight transport, which remains a key factor of regional economic development; •
more local continuous consultation - especially in the context of preparing municipal development plans - and detailed dialogue with the local players. This dialogue can be organised by the consultancy firm or design department in charge of the municipal development plan, and supported by the town centre manager and the Mobility Advisors - who will have an essential role to play, with their knowledge of the subject and the local players.
1 A national coordination team (called forum of urban distribution) was set up in 1995 in The Netherlands. It unites all kinds of private and public players.
342 Logistics systems for sustainable cities Information A better communication between shippers and carriers should be established.. The shipper has the task to communicate the authorised/recommended route to the carrier. This information should be provided in the form of a map and the names of the successive roads to be taken. In addition, the information to carriers has to be organised and encouraged by the Region at any level, for example through the channels mentioned below: (a) the setting up of information desks for carriers. An information desk will be installed before the "Ring" motorway, on the highways leading to the "Ring" and at the entries of the Brussels Capital Region city. This information desk will define and clearly explain the principle, and indicate the routes on a map of the Brussels network. Road maps showing the hierarchy of the roads and the authorised routes should be available (for free or at a cost) at these desks. One could imagine finding these information desks at petrol stations and at areas provided for the long-term parking of lorries. They should also be capable of giving precise and clear information about authorised delivery hours, available parking places for lorries, etc.; (b) a free-call telephone service for use by carriers, shippers and anyone in need of information about urban freight transport; (c) the appointment of a "Mister Freight" within the Infrastructure and Transport Administration, as a source of information. This person would have a privileged relationship with the various partners involved in urban logistics. This person should be well informed about the latest changes in regulations and technologies of urban freight transport, and be a member of a network of international experts. This person will be in charge of dialoguing with the various players and monitoring the implementation of the actions recommended in the Iris Plan and the Regional Development Plan in the field of freight transport; (d) a route database. We suggest developing a database managed by the regional administration in charge of transport policy. This database should be accessible through the Internet and contain the following information: (i) the system of de facto exceptions the Region and the nineteen municipalities will have agreed upon ; (ii) for listed activities subject to a case-by-case examination resulting in a free or imposed route: the name of the activity, the address where the activity takes place, the delivery place, and the imposed route or the mentioning of a free route. Every official in charge of issuing urban permits will have to enter the data into the database each time a permit is delivered under certain conditions; (iii) the recommended routes to reach the other (non-listed) activities, the main freight areas (industrial areas, commercial centres, harbour areas, etc.), or any other destination in the Brussels Capital Region. Even if non-listed activities do not 2 We would like to remind the reader that in our opinion non-listed activities should be able to choose their routes freely.
A study on the setting up of lorry-dedicated routes in the Brussels Capital Region
343
require a permit, care should be taken to indicate a recommended route, to improve accessibility for heavy vehicles and to preserve the quality of life for residents. This database will be accessible both for enforcement and for information to shippers and delivery persons. Maps and a route guidance information system will be added to the data. For enforcement purposes, we suggest that the carrier should print and keep in his vehicle a form which mentions - depending on the destination he will have entered in the database - if the route is mandatory or not. If the route is imposed, the form will indicate the precise route to follow. This form should be kept with the delivery and/or the consignment note stating the destination. The operation of the system will depend on the way the database is updated. The imposed routes will have to be entered before issuing a permit. The management of the database will be left to the care of the regional administration in charge of transport policy. A person in charge of freight transport (see the previous Section) has to be appointed and respected. He/she will be given the necessary means for access to the municipalities and the different players in freight transport. He/she will be the privileged contact person for regulations and actions on the territory of the Brussels Capital Region.
THE DESIGN OF ROADS AND STRUCTURES Too often the gauge (weight and dimensions) and specific characteristics of heavy goods vehicles (turning radius, blind area, time needed to cross a junction, acceleration power, etc.) are not taken into consideration when designing and constructing road infrastructure and in town planning. For instance, when devices are installed to protect residential areas, one should not forget the exceptional use by heavy vehicles such as emergency service vehicles or removal vans. In areas known for their freight traffic and where the presence of lorries is justified, their access should be guaranteed, while providing for their harmonious integration in the general traffic and in parking regulations. Last but not least, it should be noted that a lot of facilities strictly reserved to lorries - in particular delivery areas marked out on the street — prove to be ineffective, as they are being occupied by other road users whose behaviour is encouraged by a lack of enforcement and penalisation. Recommendations have been made for: (a) the planning of intersections, either with traffic signals or with a roundabout; (b) the construction of traffic calming devices (such as round- and flat-top speed humps): the highway code stipulates the technical requirements (shape and dimensions) to be met by speed control devices and humps. The dimensions of humps will be variable, depending on the type of traffic (for instance to accommodate heavy vehicles like buses or lorries); (c) the design and location of delivery areas: a better identification of the areas (choice of materials, etc.) is called for, and special attention should be paid to the urban furniture
344 Logistics systems for sustainable cities (minimum distance), the comfort of van drivers and the other road users, and the accessibility of the areas (facility for manoeuvring loading/unloading devices);
Figure 7 Examples of foot-ways suitable for the loading/unloading devices (d) the choice of materials on account of their strength and the development of that strength over time: facilities have to be designed while allowing for the passage of heavy vehicles and the main objectives that have been set (mechanical strength, standard use, possible use on heterogeneous substrate, visual differentiation, skid resistance, tyre noise, accommodation to urban situation, resistance to oil and fuel, etc.); (e) the cross section of the road: outside built-up areas the width of lanes for vehicular traffic is 3.50 m. hi built-up areas, it can be reduced to 3 m. Under such conditions, the crossing of heavy vehicles or the confrontation with other types of traffic may be tricky. Permission to overflow to the right while using the drainage channel, the central reservation or a mountable cycle track could be a solution. Like the planning of areas with dense freight traffic, the planning of residential quarters should pay attention to heavy goods vehicles that could enter the area. There is currently a growing interest in laying out roads within residential areas so as to discourage through traffic of passenger cars. Of course, such planning has an impact on heavy goods traffic. Road humps, foot-way extensions and carriageway narrowings are just as disturbing for heavy traffic as for passenger cars. Nevertheless, the whole territory (roads that are not part of predefined routes included) should remain accessible to a certain number of heavy vehicles, in particular fire engines, safety vehicles, vans for the collection of household waste, and other exceptions. Moreover, unlike passenger vehicles, heavy goods traffic rarely transit in residential areas. The phenomenon of through freight traffic is marginal and seems to be due to non-existent or inadequate direction signing. Consequently, it is recommended that residential areas should be protected without hindering the exceptional crossing of junctions by, and the circulation of, heavy traffic.
A study on the setting up of lorry-dedicated routes in the Brussels Capital Region
345
For example, at junctions where the width is reduced by foot-way extensions with small posts, we suggest replacing these by mountable foot-way extensions in well-chosen (resistant) materials.
Figure 8 Example of mountable foot-way extension Residential areas may also be protected by urban furniture. One should make sure that its variety and frequency on foot-ways and shoulders do not prevent heavy vehicles from circulating. This can be the case where urban furniture is too close or perpendicular to the road. The required lateral distance of urban furniture from the road has to be investigated in this respect. In particular, the development of plantations (often used to indicate 30 km/h areas) should be included in the reflection.
CONCLUSIONS The setting up of mandatory routes for heavy freight vehicles on the territory of the Brussels Capital region results from a firm political intention to protect residential areas against any through traffic and the environmental or noise nuisance it may cause. Specific routes leading to the areas that generate freight traffic are indicated before entering the Region. They make it possible to concentrate heavy vehicles on major roads specially designed and laid out for them. In addition, vehicles heavier than 19 t can enter a residential area only if they hold an authorisation issued in the context of the environmental permit for the served activity. This permit may impose a pre-established route on the vehicles. The setting up of these routes requires adequate means in the field of signing and enforcement. A dialogue with, and information to, the parties involved are two further prerequisites to the success of mandatory routes for heavy freight vehicles.
REFERENCES Dablanc, L. (1998). Le transport de marchandises en ville, une gestion publique entre police et services, Editions Liaisons. Gerardin, B., D. Patier, J.L. Routhier and E. Segalou (2000). Diagnostic du transport de marchandises dans une agglomeration. LET (Laboratoires d'Economie des transports), Programme National «Marchandises en Ville», Direction de la Recherche et des Affaires Scientifiques et Techniques, Ministere de l'Equipement, des transports et du Logement, ADEME-CERTU- Lyon. CERTU - ADEME (1998). Plans de deplacements urbains. Prise en compte du transport des marchandises, ADEME-CERTU, Lyon.
346
Logistics systems for sustainable cities
CERTU - ADEME (1999). Plans de deplacements urbains. Prise en compte de la pollution de l'air, du bruit, et de la consommation d'energie, Guide methodologique, ADEMECERTU, Lyon. CERTU (2000). Dimensionnement des structures des chaussees urbaines. Methodologie de conception d'un catalogue adapte au contexte local, CERTU, Lyon. CERTU (juin 2001). Transports exceptionnels et amenagements de voirie en milieu urbain, CERTU, Lyon. CERTU, ADEME, Ministere de l'Equipement, des Transports et du Logement, (2002). Plans de Deplacements Urbains et Marchandises en ville. Reflexions a destination des elus, Dossier 129, ADEME-CERTU Lyon. CERTU, Ministere de l'Equipement, des Transports et du Logement. (2002). Logistique urbaine en Europe. Quelques elements statistiques et experience de regulation dans des villes europeennes, CERTU, Lyon. CERTU, (decembre 2002). Structures et revetements des espaces publics. Guide technique. CERTU, Lyon. ECMT (1997). Les transports de marchandises en ville. Rapport de la 1094me table ronde d'economie des transports, ECMT-OECD, Paris. European Commission, Directorate General Transport (1998). COST 32: Urban Goods transport. Transport Research, Final Report of the Action, European Commission Luxembourg. GART (Janvier 2001). Les collectivites territoriales et le transport de marchandises. Compte rendu des reunions du groupe de travail « Marchandises » du GART en 19992000, GART. GART, FNTR, (fevrier 2000). Mieux gerer les marchandises en ville. Guide d'actions, GART, Paris. IBSR: Institut beige pour la Securite Routiere, legislation routiere, IBSR, Bruxelles. LET: Laboratoire d'Economie des Transports (octobre 2000). L'integration des marchandises dans les systemes de deplacements urbains. Treiziemes entretiens Jacques Cartier, ISH-ENTPE. Ministere de l'Equipement et des Deplacements de la RBC, Service de la Politique des Deplacements, (1997). Plan de Transport Marchandises de la Region de BruxellesCapitale, Bruxelles. Ministerie van de Vlaamse Gemeenschap, Departement Leefmilieu en Infrastructuur, Administratie Wegen en Verkeer, Afdeling Verkeerskunde (augustus 1997). Vademecum Rotondes. RTS : Recherche - Transports - Securite, n° 69 (octobre-decembre 2000). « Special Plans de deplacements urbains ». RTS. Tritel, Iris Consulting (december 1999). Begeleiding van de Administratie Wegen en Verkeer bij de implementatie van de wegencategorisering en het opstellen van ontwerprichtlijnen. Eindrapport. Vallar, J.-P. (2000). Gestion du trafic commercial de livraison. Actions de villes europeennes. Etat de l'art 1997. ADEME, Paris, van Binsbergen A. and J. Visser (May 2001). Innovation steps towards efficient goods distribution systems for urban areas, The Netherlands TRAIL Research School. Internet Sites EU: EU:
BESTUFS: www.bestufs.net ELCIDIS: www.elcidis.org
A study on the setting up of lorry-dedicated routes in the Brussels Capital Region 347 EU: EU: NL: F:
COST 321: transport de marchandises en ville : www.cordis.lu/cost-transport CEMT (Conference Europeenne des Ministres du Transport): www.oecd.org/cem/ Platform Stedelijke Distributie: www.psd-online.nl www.gart.org, www.certu.fr, www.transports-marchandises-en-ville.org, www.predit.prd.fr.
This page is intentionally left blank
25
NEW CONCEPTS FOR CITY LOGISTICS Prof.ir. Joan C. Rijsenbrij, Delft University of Technology, the Netherlands
ABSTRACT The continuing urbanization, the still increasing cargo distribution volumes towards the city centers and the spectacular growth in road transportation are deteriorating the traffic situation in large cities, and so a heavy threat for fast, predictable and cost-effective city logistics. Now is the time to develop new standardized logistic concepts, including the use of all available modes of transport. Standardized city boxes, vehicles, handling equipment and city access rules will support the interchangeability within logistic networks and will encourage the implementation of new concepts for city logistics offering a better service for equal or even lower costs. Shuttle concepts for road-bound distribution will result in better vehicle utilization, less environmental impact and better quality of life. Furthermore, rail-bound systems for the combined transport of passengers and freight have a large potential for regional and city-bound cargo distribution. For area's with a suitable waterways network, barge transportation can be used both for longhaul and bi-modal (barge-truck) city logistics. Partnerships between transportation companies, retailers, logistic providers and local/regional authorities will be required to introduce new logistic concepts. These will be necessary to maintain efficient city logistics and to improve the quality of life in the cities.
INTRODUCTION For at least 100 years, city logistics has been used to improve the quality of life in larger cities. For many centuries, the distribution of goods was done by means of persons, animals, barges and carriages (carts). In many flourishing cities, impressive infrastructure consisting of a network of canals and paved avenues supported the daily supplies of consumer goods and building materials, sometimes for freight only but regularly for combined passenger and cargo transportation as well. The long-haul stretches; both nationally and internationally (coastal)
350
Logistics systems for sustainable cities
were often maintained with combined transport, using sail boats, tow boats and mail-coaches (Figure 1).
Figure 1 Tow barge for combined transport The development of rail transportation continued this combined transport but in the 20th century the development of trucking and metro systems induced an almost complete separation between passengers and cargo. Nowadays trucking is the dominant mode for the distribution of goods in cities and there is an increasing conflict between the mobility of people (trams, metro, cars and bicycles) and the distribution of cargo. However, there is a growing awareness about the need for efficient city logistics, hence the reliable distribution of goods is essential for the vitality of cities. This paper focuses on recently developed new concepts for city logistics and one of the most important critical success factors: standardisation. New concepts have been developed for multi-modal operations and will contribute to a better utilisation of infrastructure and a better environment in the cities. However, a successful implementation requires a certain level of standardisation both for the transport units (city boxes) the load area's of vehicles, handling techniques, information control and management of logistics.
KEY ISSUES IN CITY LOGISTICS Globalisation, world-wide communication systems and ongoing urbanisation has changed city life considerably over the last 5 decades and so has city logistics. More than 40% of the world population is concentrated in cities and this percentage will increase. By the year 2050 almost 4.5 billion people will live and work in and around cities (some of them with more than 15 million people) and this will cause tremendous supply demands. The following key-issues for city logistics and the requirements for transportation systems can be recognised: (a) increasing volume of goods moving into and within cities with smaller drop-sizes are getting smaller, last minute deliveries are increasing; (b) offices, retail outlets (supermarkets and department stores) and catering (restaurants, bars, entertainment) are the dominant branches demanding daily supplies; (c) there is an increasing amount of DINKI's (Double Income No Kids) and senior people who request home deliveries for its convenience. Small companies as well ask for daily supplies to be delivered (catering and stationery). This has caused an increase in the number of companies, specialising in delivery services;
New concepts for city logistics
351
(d) care for the quality of life has increasingly resulted in the realisation of reverse logistics (waste material, recyclable products, household appliances, glass and chemicals). This generates additional material flow, often starting at shops and supermarkets; (e) E-business results in many small parcels (CD's, books, toys, electronics and software) to be delivered at homes and offices daily; (f) municipal directives require the reduction of CO2 , NOx and noise pollution which is supported by restrictions in permitted vehicle dimensions (height, width, axle loads and GVW) and requirements for special drive systems (LPG and electric); (g) the growth in traffic congestion and the establishment of various traffic regulations (circulation plans, access time slots and reduced speed zones) have resulted in decreased average speeds of city delivery trucks. In some cities the average delivery speed has dropped to approximately 10 km/h.; (h) there is a growing amount of partnerships and alliances between logistic providers, including (privatised) mail and parcel service companies; (i) the realisation of new infrastructure is not related to traffic growth and society is asking for better utilisation of existing facilities; (j) there is a fundamental mismatch in volumetric cargo mass for long distance vehicles and city distribution trucks (Figure 2); (k) a growing concern about working conditions for labour in the distribution industry is stimulating the use of roll-containers and automated loading/discharge provisions. The retail industry and parcel services are increasingly changing towards rolling cargo.
Figure 2 Mismatch in volumetric masses for long haul and city trucks The above trends should be considered when managing existing and developing future transport systems. City administrations are increasingly aware of the importance of a reliable, amply available cargo transportation, which is necessary to maintain the vital functions of the city.
352
Logistics systems for sustainable cities
STANDARDIZATION, A CRITICAL SUCCES FACTOR During the last decades many new concepts for city distribution have been developed and some of them were even put into operation more or less successfully (Figure 3). Probably the absence of sufficient critical mass and inter-changeability hampered a real break through for any of these concepts. Besides, the application of special designed vehicles and load units in small series caused rather high operating costs.
Figure 3 Examples of concepts for city distribution Nevertheless, the industry has shown many examples of successfully introduced new concepts for transportation, often based on standardised, mechanical equipment and standard work procedures. Standardisation is probably the most important pre-requisite for fast, safe and efficient handling and transportation of cargo. Examples are the king-pin of semi-trailers, pallet dimensions, unit load measurements (uni-cube) and the world-wide spread of maritime containers (with approximately 15 million TEU of boxes world wide). The tremendous success of containerisation can be attributed to some of the following factors: (a) the rigid standardisation resulted in simple vehicles and handling systems; (b) the only regulations for containers regarded the 8 defined corner points in space, the shape and dimensions of the corner fittings and maximum lifting loads; (c) a variety of container types both for general cargo and special commodities (refrigerated cargo, liquids, bulk, over height cargo, etc.) were developed to suit particular niches in the cargo trade; (d) substantial savings could be achieved in all parts of the transportation chain (due to increased scales, competition and efficient methods). The cost component of the container itself (in the overall cost) was reduced to less than 5% of the total transportation cost; (e) the containers offer a good protection (weather and pilferage) and can be used for temporary storage of goods;
New concepts for city logistics
353
(f) each container is given an unique identification number and size-type code, which facilitates "tracking and tracing" and planning, important conditions for mechanisation and automation. The arrival of various small distribution boxes and the increased use of rolling cargo has encouraged research towards identifying a universal load unit for distribution services. The objective was to develop standardised characteristics, allowing for a simple design and multiuser application. This research resulted in a set of functional requirements for a city box. The most important ones are: (a) outside dimensions for mass-volume road haulage and convenient city movements: 255 x 215 cm (road: 255 width; city: 215 width); (b) internal height comfortably for human access: 210 cm.; (c) pay-load about 3000 kg; (d) the bottom should have all functionality for handling ( forks, rollers, skidding), fixations and box stiffness; (e) the box must be identifiable with reliable, using simple techniques; (f) the box should allow an easy, low-noise loading and discharging of roll-containers; (g) the box must be aesthetic, easy to clean and effectively lockable; (h) the box cover should allow a variety of applications for various branches. A first prototype of such a box (Figure 4) has shown that the application of aluminium and sandwich panels of synthetic material can result in a tare mass of about 300 kg. Lashing methods from the air cargo industry will allow for simple effective means of fixation onto vehicles.
Figure 4 City Box A standardised box enables the development of a variety of other facilities required for an efficient distribution of goods into cities. Multi-modal transportation demands for standard provisions on vehicles, barges, trains, metro's and the related warehousing functions in district centres, logistic parks, retail outlets and shopping centres, etc. However, the standard city box bottom allows for simple modifications onto existing equipment. The selection of one (or a few) techniques(s) for the identification of city boxes will facilitate tracking and tracing. Transponder and bar code techniques are already widely and reliably in use within the warehousing and retail industry. Standardisation of city boxes, vehicles and handling equipment will definitely encourage the inter-change-ability within logistics networks. "Grey boxes" will be helpful to avoid empty legs for unbalanced supply chains and in addition standard city boxes for many supply
354
Logistics systems for sustainable cities
branches could be transported by common carriers serving these various branches with optimised (cost effective) transportation networks. Local and regional authorities as well should standardise their rules for access into city centres. This should not be limited to vehicle mass and dimensions but should include measurements for time windows and traffic corridors as well. New sustainable, environmental friendly concepts could be supported by allowing such concepts the use of existing, congestion free traffic lanes for public transport (tramway, buses, etc.). The success of the maritime container and many examples in the manufacturing and IT industry show the tremendous savings that can be obtained with standardisation, although it requires compromises from the individual user. However, for city logistics the following statement will also be valid: Standardisation: nobody's best, for everybody better.
NEW DISTRIBUTION CONCEPTS The introduction of one standardised city box platform (with many branch-specific designtypes) will result in many new supply/distribution/collection concepts both in single mode and multi-modal transportation systems. The differing conditions for long-haul transportation and the distribution of cargo in the cities demand differing means of transport. The long-haul is focussed on speed, costs and high capacity (economy of scales); the city-leg emphasises safety, quality of life, manoeuvrability and fast handling. A de-coupling of long-haul and city-leg (in logistic centres or a simple interchange area at the city border) sometimes with a small buffer capacity will support cost-effective transportation. The benefits of dedicated transportation systems for each trajectory will more than compensate for the disadvantages of an interchange. Recently the following new concepts have been announced: (a) shuttle concepts for road-bound distribution; (b) rail-bound systems for the combined transport of passengers and freight; (c) waterborne concepts, both for long-haul and city distribution. These new concepts are further detailed below. Shuttle concept for road-bound distribution This concept is based on splitting-up the logistic chain between distribution centres and the branches (retail outlets) in two parts: a supply part on the main road and a distribution part in the city. Both parts are connected through an interchange at the edge of the city or city centre. The shuttles transport the cargo in city boxes by means of large road trucks (semi-trailer type) between the distribution centre and the interchange area, and this type of transportation is highly efficient (max. truck utilisation and application of cost effective large trucks) and results in the lowest environmental impact per ton kilometre and the highest infrastructure utilisation (minimum amount of trucks per cargo shipment).
New concepts for city logistics
355
At the interchange location the city boxes are positioned on city delivery vehicles and during that operation the boxes are turned 90 degrees resulting in a smaller width for the city distribution leg. For the delivery process a family of vehicles can be applied. For the efficient delivery of larger drop sizes a 2-box delivery truck is available (Figure 5). This vehicle is about 7.4 m. long, 2.25 m wide and 3.2 m. high with an GVW of maximum 11 tons. This vehicle is based on a standard small truck featuring low noise levels and low emission (Euro class 4 and in the future Euro class 5) and many provisions for safe and comfortable driving in cities.
Figure 5 Two box delivery truck and three box multi trailer train For pedestrian area's and/or shopping centres a multi-trailer train with 3 city boxes, towed on small steerable carts behind one tractor, can be applied (Figure 5). This medium speed trailertrain (30 km/hr.) can deliver the goods to one outlet or various outlets, even for various branches. For smaller drop sizes and door-to-door delivery for parcel services a one city box truck will be an attractive vehicle. The interchange area at the city border (or in the city near the city centre) can be very simple such as a loading platform possibly connected to a small automated warehouse for 10-30 city boxes. A low-cost and very flexible interchange between the long-distance shuttle trucks and a 2 box delivery truck can be realised through transfer mechanisms with roller beds on both trucks (Figure 6).
356
Logistics systems for sustainable cities
Figure 6 City box interchange from large truck to delivery truck In a case study the feasibility of the above new concept was proved, using actual delivery demand data from 10-30 supermarkets requesting for a daily delivery service within a one-hour time slot and with a service level of 99%. In the case study the net present value method was used for an economic evaluation and this showed that the split operation (long-haul shuttle + city delivery) may provide a supply chain cost reduction of 5-10% (depending on supply profile and geography). On top of that a reduction in environmental impact (15-30%) and a reduction in infrastructure utilisation (vehicles per km) of 10-25% will be achieved. This new transportation concept for city logistics can be further enhanced by municipalities when the special city distribution vehicles are allowed on the (now under utilised) bus, tram and taxi lanes, which will create further cost savings and service improvements.
Rail bound systems for the combined transport of passengers and freight Passenger (light rail) trains, tramways and metro systems must fulfil some rigid demands: a frequent service with predictable/reliable arrival times, and these requirements make them attractive for city logistics as well (just like in the early days). The dedicated (non-disturbed) infrastructure for light-rail and metro systems supports a high service potential, although substantial modifications in stations will be required. Metro systems for urban freight transportation will have to fulfil some major requirements: (a) the average speed should be maintained at 40 km/hr and the time stopped at stations should be less than 30 seconds; (b) cargo flows and passenger flows should be completely separated (cargo security and the potential for automated cargo handling); (c) there should be an automated cargo identification to allow an easy tracking and tracing and efficient process control; (d) connecting transportation to the final consumer should have very good (integrated) access to the station with a high frequency delivery/receival service; (e) some buffer capacity (e.g. 1-2 hours capacity) will be an advantage to realise an attractive metro car utilisation and an effective distribution system.
New concepts for city logistics
357
Freight transportation could be realised by special freight train sets (3-5 cars) or by metro trainsets composed of 3-6 passenger cars and one freight car. The latter (combi-train) alternative will offer the best service level both for passengers and freight. Figure 7 A and B shows two basic alternatives from which alternative A has the lowest complexity both for city boxes (no top lifting provisions required) and the handling facilities. However the train sets must be stopped with a position accuracy of ± 5 cm.
Concept B. Figure 7 Alternative concepts for combined passenger and freight transport The high hourly service frequency of metro (light-rail) systems include a substantial city transport capacity which can be demonstrated with an imaginary metro line, connecting the city centre with offices and shopping area's and suburban area's (Figure 8). In a case study for a city with 5 metro lines of 20 kms each and with 10-20 trains per hour, a carrying capacity for 8 city boxes and an average pay load factor of 50%, the transportation capacity was 324.000 tonkms/day. This is equal to the transport capacity of more than 425 delivery trucks and so, city logistics could benefit from a built-in surplus capacity in metro and light-rail systems.
358 Logistics systems for sustainable cities
Figure 8 Imaginary metro line Waterborne concepts, both for long-haul and city distribution Since thousands of years B.C., large and heavy cargo was only transported by ship and that continued until the end of the 19th century. For centuries, sailing barges and tow-barges were the most important means of transport for inland distribution. Cities were often established along waterways and in many large trading cities extensive water networks were created. In the 20th century rail transportation and especially road trucking replaced water-borne transportation. However, the increasing traffic congestion have supported a revival of barge transportation, even for fast moving consumer goods and other general supplies for city bound activities. In some Dutch cities a few specific applications have proved to be successful both from a service and a cost point of view. Parcel services (DHL, Amsterdam) beer and beverages distribution to bar's and restaurants (Utrecht) and general cargo distribution (Venice) are only a few examples (Figure 9).
Figure 9 Successful applications in barge transportation A further developed and more generic application is a multi-modal concept for the transport of city boxes by means of dedicated barges and special delivery trucks for the final delivery to shops, offices or locker-stations (for E-commerce activities to households). With this concept barges transport city boxes from one (or a few) distribution centres in the city towards a number of interchange locations at the water network (Figure 10). For cost reasons (and
New concepts for city logistics
359
maintainability) the barge is provided with a lifting device (either a crane or a lift-able rollerbed) to handle the city boxes onto (from) the shore. A similar concept is now under development for the city of Venice where standard VenIX boxes will be transported between a distribution centre and drop-off locations.
Figure 10 Barge - shore / vehicle interchange area Barges are even more attractive for long-haul transport when sufficient cargo volume can be attracted. Low cost, reliability and a predictable service are the major advantages, which made a number of large shippers decide to start an inland waterway-shipping network for retail logistics. Palletised commodity cargo (beer, beverages, etc.) is shipped in barges, which serve as a floating warehouse (Figure 11) connecting factories with distribution centres of retail chains and in the future with city distribution centres. The onboard (semi)-automatic pallet rackings (up to 1000 pallets) and pallet handling equipment make these self-sustained barges very attractive for cross-docking.
Figure 11 Barge "River Hopper" for palletised commodity goods The largest advantage of water-borne transport for city logistics is its limited demand for infrastructure. Moreover once the waterways network is installed (existing) there is a tremendous transport capacity and high circulation reliability.
360
Logistics systems for sustainable cities
ASSESSMENT OF NEW CONCEPTS The feasibility of new concepts must be checked, both their logistic service, environmental impact and economics. The logistic performance (transport capacity, delivery reliability, flexibility for stochastic influences, etc.) can be analysed with the help of a dynamic simulation tool, with object-oriented modelling using special simulation packages like Tomas-Delphi or Em-Plant. Such a model represents the logistics concept including all major components and their characteristics: e.g. the distribution centre, an interchange area, the various transport trucks, the retail outlets and their demand specifications. For the earlier described trucking shuttle concept such a model has been developed and the model was loaded with actual data about destinations, drop-sizes, transportation times and above all, the stochastic variation of driving times and handling processes. Important elements in the simulation model are the planning and control modules that have been modelled according real-life logistic controls. The model allowed a comparison of various concepts (truck sizes, delivery patterns, etc.) and the analysis of future traffic characteristics on the logistic performance of new concepts (longer travel times and more unforeseen traffic jams). On top of that the model generates the travel times and travel distances per type of equipment and registers the logistic performance (service levels, including the average, 95% values and extremes). Experiments with the model showed that the splitting-up the supply chains, in order to benefit from the typical truck characteristics for long-haul (large scale) and city (small and environmental friendly) does not influence the delivery service level. To the contrary, the additional interchange allows a higher utilisation rate of the large shuttle equipment and a small buffer capacity at the interchange improves the robustness of the logistics system (less vulnerable to stochastic disturbances). Moreover it was shown that the new concept performs better when the distribution centre is located further away from the city centre(s). The coupling of the simulation results (kilometres driven and driving times) to an economic evaluation tool (NPV-method) showed that a split-up operation with 2-box delivery trucks or 3-box multitrailer trains will be less costly than the present single trip operations with tractor/semi-trailers.
ACTIVITIES TOWARDS A SUCCESSFUL IMPLEMENTATION The new concepts described and requirements for standardisation have the potential for improved (more reliable and less costly) distribution of goods in the cities with an improved quality of life. The introduction of such concepts will demand the following activities: (a) identification of cargo flows requiring daily services to a variety of city destinations (office supplies, fashion and department stores) under the control of large logistic organisations ( 3PL-providers, retailers, parcel services and transportation companies); (b) structuring of logistics chain with multi-modal supply links to city logistics parks possibly supported with one or a few city distribution centres (Figure 12). From there an intricate structure for distribution activities can be arranged;
New concepts for city logistics
361
Figure 12 Structuring city distribution (c) the introduction of a single standardised city box. City boxes should have generally accepted basic characteristics such as main dimensions, provisions for handling, reliably readable ID-number, simple means for loading and unloading the box, interchange-ability acceptable for box pools, etc.; (d) the development of dedicated vehicles with minimal standardised requirements. Designs for trucks, barges, railcars, metro cars, etc., should follow the developments of the related industry in order to avoid being too complicated equipment and too expensive to maintain; (e) the conducting pilot studies, to gain experience, to stimulate acceptance and to show the potential of new multi-modal city distribution concepts. For these pilot studies, there should be commitment from a number of users, who will contribute with a base load of freight transportation. Such "founding fathers" can be large department stores, major office supply companies, city logistics providers or city distribution companies. A good example is shown in Dresden where a major industrial company (VW) supported a special freight tramcar in a public passenger transport system (Figure 13);
Figure 13 Freight tramcar in Dresden (f) the interface between the long-haul stretches and the final city delivery legs (i.e. citybox cross-docking facilities, metro stations or barge-truck interchanges) should be designed for reliable operations and might even include some buffer capacity. The state-of-the-art in automated high-rise warehouses and air freight transportation systems will support the design of such reliable interfaces; (g) the development of efficient process control systems with proper interfaces for order acceptance/confirmation and intelligent modules for capacity planning (optimising available capacity and agreed service levels). A proper tracking and tracing system, accessible through a web-side should allow operators and users an on-line information about the status of their shipments.
362
Logistics systems for sustainable cities
The variety of goods for various branches and the many different types of supply chains will thwart a fast introduction and wide acceptance of new improved concepts for city logistics. However, authorities and key players in city logistics should co-operate in order to install more standardised, interchange-able multi-modal transport systems. That's the only way to guarantee adequate supply systems with improved quality of life in cities at a reasonable cost.
CONCLUSIONS Continuing urbanisation and increasing material wealth will result in higher demands for the transportation of goods within large cities. The spectacular growth in road transportation has resulted in severe traffic congestion, especially in the city centres and so, there is an increasing concern about the quality of future delivery systems for consumer goods (reliability and costeffectiveness). Furthermore, there is growing awareness concerning quality of life issues in cities and the limitations of expanding transportation infrastructure within cities. This will require higher utilisation of existing city transportation systems including better use of existing infrastructure such as roads, rail trucks, and waterways. The conditions for urban cargo distribution have changed drastically over recent decades and now the time is right to develop new standardised concepts, including the use of alternative modes of transport. Waterborne systems and rail systems have a large potential and can be combined with road-trucking systems, composed of efficient large shuttle trucks and cityfriendly delivery trucks or rubber-mounted multi-trailer trains. A split-up of distribution processes will result in cost savings and a better quality of life in cities, however a prerequisite is the application of standardised city boxes that should be launched and supported by large actors in the industry in order to create a "de facto" standard. The deteriorating traffic situation in large cities and the tremendous cost and time consuming procedures to realise new infrastructure in the city centres are major limitations in the transportation of consumer goods. If society demands more and faster delivery of goods and at the same time a better quality of life in cities than something must be done. Therefore partnerships between transportation companies, retailers, logistic providers and local/regional authorities should be formed to introduce new logistic concepts that will result in a win-win situation for all parties involved. The new concepts presented can make a contribution. The cargo is there, the technology is available and operational organisations can be arranged. It is the sense of urgency, the belief in the innovative approaches themselves, and the strength of partnerships that will determine whether city logistics will be improved or not in the future.
REFERENCES Binsbergen A.J., A.J. Klein Breteler, J.W. Konings, J.C. Rijsenbrij, J. Katgerman and D. Piebenga (1998). Continentale laadeenheden voor intermodaal vervoer, CTT publicatiereeks 35, Delft. Binsbergen van A.J. and J.G.S.N. Visser (2001). Innovation Steps Towards Efficient Goods distribution Systems for Urban Areas, Trail Delft.
New concepts for city logistics
363
Boerema, R. and J. Vermunt (2002). DISTRI-SHIP, inland waterway shipping for retail logistics. Publication Holland International Distribution Council, Zoetermeer, the Netherlands. DHV Milieu en Infrastructuur (2001)., Met je goeie goed in de metro, DHV, Amersfoort. Hansen I.A., A.F. Kort and P.B.L. Wiggenraad (1999). Transport Operation and Management Volume A: Public transport, TU Delft Civil Engineering course CTvk4810, Delft. IVV Amsterdam (2002). Conceptnota Goederenvervoer Amsterdam. Gemeente Amsterdam dienst Infrastructuur Verkeer en Vervoer. Luiting Maten, J.H.D. (2001).Verkenning mogelijkheden multimodale goederdistributie centra in Amsterdam, TU Delft Faculteit OCP - Transport en Logistieke Techniek Rapportnummer 2001.TT.5520 Delft. Pielage, B.A. (2000). Subterranean Inland Transportation. ICHCA 2000 conference, Rotterdam. Pielage, B.A. (2000). Design approach and prototyping of automated underground freight transportation systems in the Netherlands ISUFT 2000, Delft. Rijsenbrij, J.C. (1999). Balanced scales!? Inaugural speech Delft University of Technology, Delft. Rijsenbrij, J.C. (1994). Automation: a process redesign. Europe Combined Terminals B.V., Rotterdam.
This page is intentionally left blank
26
URBAN RAIL AND INTERMODAL FREIGHT STRATEGIES IN THE ZURICH AREA: A CASE STUDY FROM SWITZERLAND
Martin Ruesch, Rapp Trans Ltd., Zurich, Switzerland
ABSTRACT This contribution describes the urban freight structure in the Zurich area, the possibilites and the potential for rail and intermodal transport. It contains results from studies made for the authorities of the city and the canton of Zurich and also from European research projects. It also notes supporting measures and framework conditions to reach a modal shift from road to rail based transport.
INTRODUCTION AND PURPOSE OF THE WORK The conurbation of Zurich is the most important metropolitan area within the urban network of Switzerland with about 1.1 million inhabitants and 600,000 employees (Figure 1). Zurich is the economic capital of Switzerland with high importance in services within the financial sector, insurance sector, consulting, technology and tourism. The conurbation of Zurich has been changing within the last 30 years into a typical consumer region because the industrial production has been shifting to neighbouring regions and was replaced by services as well.
366
Logistics systems for sustainable cities
Figure 1 Location and urban area of Zurich As other conurbations in Europe, Asia and the United States the urban area of Zurich faces increasing problems relating to urban goods movements as the growing freight traffic, lack of suitable infrastructure for deliveries, such as conflicts with other road users, including pedestrians within the historic centres, pollution and noise emissions etc. (Glucker and Ruesch, 2001). The growth in urban goods movements is mainly caused by developments and trends in the logistics and freight transport market, from individualisation of freight demand and globalisation and internationalisation of the procurement and sales markets (Ruesch and Eugster, 2001). These developments lead to decreasing consignment sizes and an increasing number of deliveries. The identified problems have negative impacts on the quality of life and the location attractiveness of the urban area of Zurich. Therefore the government of the Canton and City of Zurich initiated and supported different activities and projects on a local and regional level to achieve a more sustainable goods transport system. These activities and projects include strategies and measures in urban goods transport, such as site evaluation for terminals, supporting measures for rail and intermodal transport, use of private sidings as transhipment points, co-operation/city logistics, compact city terminals and low emission lorries. This paper summarises the results of the freight demand analysis and the projects on the role of rail and intermodal transport within urban distribution.
RELEVANT URBAN FREIGHT PROBLEMS Within the urban area of Zurich the following relevant problems relating to urban freight have been identified (Ruesch and Eugster 2001; Glucker and Ruesch, 2001, and Ruesch and Ruggli, 1998): • Enormous increase of vans for city distribution with increase of number of trips and road mileage (with conflicts as second row parking on streets, high space use, etc.) • Lack of suitable infrastructure for deliveries (eg. ramps, areas for loading/unloading, reserved parking spaces) • Critical access for goods vehicles to pedestrian zones and historic centres • Conflicts with other road users during delivery operations (loading and unloading) • Negative environmental effects such as noise and air pollution (high emissions per tkm)
1 The growth rate of the urban freight transport (relating to ton-kilometres and vehicle-kilometres) was in former years much higher than in passenger transport (passenger-kilometres).
Urban rail and intermodal freight strategies in the Zurich area
367
•
Low transport efficiency (high costs per tkm) because of small consignments and time delays in congestion • Lack of co-operation and co-ordination between transport companies • Decreasing share of rail freight transport and high pressure on urban railway goods infrastructure (urban goods stations and private sidings). The city inquiry within the BESTUFS project2 (Gllicker and Ruesch, 2001) showed that the urban region of Zurich has similar problems as other medium sized cities in Europe.
FREIGHT TRANSPORT STRUCTURE AND DEVELOPMENT IN THE ZURICH AREA Within the project relating to the strategic development of freight transport in the canton of Zurich (Ruesch and Eugster, 2001) and about freight transport within the conurbation of Zurich (Ruesch et al 2003) the road and rail freight transport data between 1991 and 2001 was analysed including volumes, commodities, origin-destination flows, vehicle parking and transport distances. Comparable data for road and rail was available for 1993 and 1998. For the data preparation, evaluation and presentation an access data base and GIS were used.
Actual situation and former development Between 1993 and 1998 the freight transport volume in the region of Zurich grew from approx. 74 Mio. tons per year towards approx. 82 Mio. tons per year (+ 11%, Figure 2).
Figure 2 Freight volumes by mode 1993 and 1998 (Ruesch et al 2003) The increase in international transport (+25%) was much higher than in Swiss-related transport (around 10%). The main reason for this is the ongoing globalisation and internationalisation of freight transport which leads to more freight flows between the Zurich region and other European regions. Whereas the share of pure rail transport slightly decreased from 10.5 to
2
BESTUFS: Best Urban Freight Solutions, Thematic network project within the 5* framework program of the European Union and also cofinanced by the Swiss Federal Office of Education and Science.
368 Logistics systems for sustainable cities 10.0%, while the share of intermodal transport grew from 1 to 1.5%. One main reason for this development is the ongoing decline of industrial activities for the benefit of the services sector. This leads to less railway compatible freight commodities (less bulk and smaller consignments). When we take a closer look at the volumes by sub-region we can differentiate between the more services oriented and more industry oriented sub-regions (Figure 3). Within the characteristic consumer regions the inbound freight traffic is much higher than the outbound traffic. The share of rail and intermodal transport depends strongly on the specific location of production plants, distribution platforms and storage facilities within the sub-regions.
figure i freight volumes by mode and sub-region iyy»(in 1OUU t per year, Kuesch et al 21XUJ The data analysis also showed that more than 50% of the freight volume is transported within the Zurich region on distances below 40 km (Figure 4). Over such distances road transport dominates and due to economic reasons there is no potential for a mode shift.
Figure 4 Freight volumes related to the region 1998 (in lOOOt per year, Ruesch et al 2003) Road Transport dominates freight transport with a share of about 88% (in 1998, Ruesch et al 2003). Rail freight transport has approximately 10% and intermodal transport rail/road has a
Urban rail and intermodal freight strategies in the Zurich area
369
share of approximately 2%. Rail's share is constantly decreasing, while intermodal transport is increasing. Reasons for this development are the changes in commodity groups (from bulk-to consumer goods), consignment sizes and changing requirements relating to lead times. Also, increasing containerisation of goods can be observed. From the analysis by the transported commodity groups (NST/R-classification) we can make the following statements (Figure 5): • Sub-regions with a share of consumer goods (food and non-food) higher than 66% are the City of Zurich, the City of Winterthur, Glattal Siid and Limmattal. Most of the shopping centres and restaurants are situated within these regions. • Sub-regions with a high share of construction site related commodities such as gravel, cement, excavated material are Zurcher Unterland and Weinland (high number of gravel pits and cement production plants). • Sub-regions with a high share of mineral oil are the Glattal Nord and Zurcher Unterland (tank storages and airport). • The commodity groups of iron, steel and ore are of minor importance and reach only in the sub-regions Limmattal and Baden substantial shares (industrial areas). • Chemicals are in all of the sub-regions of minor importance except in Brugg Zurzach. When we take a closer look at the road freight transport we can see that 95% of the freight volumes (in tons) are transported by heavy vehicles (>3.5t) and only 5% of the volumes are transported by lorries and vans below 3.5t (Ruesch et al. 2003, Figure 6). But the vans and lorries of less than 3.5t generate in the Zurich area about 60% of the mileage (Rapp and EBP, 1998). Vans and lorries below 3.5t have a substantial share in the commodity groups such as vehicles, machinery, semi-finished products including courier services.
Figure 5 Freight volumes by commodity group 1998 (in lOOOt per year, Ruesch et al 2003) Overall, 65% of the rail and intermodal freight volume within the inner sub-regions was transported via only 3 (12%) out of 23 public goods stations (Figure 7). The 16 public goods stations had only a share below 2% each. Around 60% of the volumes are transported via private sidings and 40% via railway goods stations. With 75% being inbound traffic and only 25% outbound traffic, the loading factor of the rail wagons was consequently low.
370 Logistics systems for sustainable cities
Figure 6 Road freight transport by commodity group and vehicle size 1998 (Ruesch et al 2003) Rail volumes are increasing slightly but the share of rail transport is decreasing at the same time because of the fast growing road transport segment. During the last 10 to 15 years the railway company concentrated the rail traffic with a reduced number of public goods stations.
Figure 7 Rail freight transport by railway goods stations 1998 (in lOOOt per year, Ruesch et al 2003)
Driving factors for freight transport development and forecast 2015 and 2025 The main driving factors for the former and future developments of urban freight have been identified (Ruesch et al 2003): . the economic development with changes from the industrial to the services sector and an increasing income of inhabitants including increasing consumer goods flows • the individualization of demand including increasing just in time transport, decreasing consignment sizes and increasing deliveries • the globalisation and internationalisation of the procurement and sales markets including increasing international traffic, increasing containerisation of transports and longer transport distances • the concentration on key activities including outsourcing of non-core activities, logistics and transport tasks, increasing transport demand for semi-finished products • strategic co-operation and formation of freight integrators with bundling possibilities and increasing transport efficiency
Urban rail and intermodal freight strategies in the Zurich area • •
371
use of new technologies (eg. information and communication and e-commerce) including improving information flows and decreasing consignment sizes concentration of logistics activities within logistics regions and zones with synergies between logistics and transport actors and increasing transport distances.
An important influence has been the trends in the framework conditions. This includes the liberalisation and harmonisation of the transport markets, the enlargement of the European Union, the growing capacity problems on the transport networks and the introduction of road pricing (eg. Heavy Vehicles Fee in Switzerland). The future transport volumes by mode have been estimated (Ruesch et al. 2003, Figure 8). The chosen approach considered various factors such as the trends in logistics and economic development for the development scenarios for 2015 and 2025. For long term urban planning issues rough estimations were suitable with an identification of the possible range of future volumes.
Figure 8 Freight transport forecast and scenarios (Ruesch et al 2003) For the year 2025 freight volumes between 75 Mio. t and 154 Mio. t are expected depending on the investigated scenario. Due to the numerous uncertainties, the range is quite large. The strongest increase is expected in international transport. The share of rail is expected to be between 8 and 15% and the share of combined transport between 3% and 6%. In addition the potential for modal shift from road to rail (including intermodal transport) for the inner region of 3 to 4 million t has been estimated taking into account different shares of rail relating to the distance ranges.
OVERALL URBAN FREIGHT TRANSPORT STRATEGY Based on the freight transport developments and the raised problems, the city (Stadtplanungsamt Zurich, 1995) and the Canton of Zurich (Amt fur Verkehr Kanton Zurich, 2001) defined a number of objectives and measures for freight transport (Table 1).
372
Logistics systems for sustainable cities
Table 1 Regional and local freight transport strategy (Stadtplanungsamt Zurich, 1995 and Amt fur Verkehr Kanton Zurich, 2001) Main Objectives
Main measures / instruments
Canton Zurich • Provision of high quality transport network • Excellent accessibility of the Zurich area • Improving transport safety • Reduction of negative impacts on the environment, urban areas and landscapes
City of Zurich • Optimisation of freight transport chains • Optimisation of use of freight related infrastructure • Improving efficiency of freight transport • Reduction of negative environmental effects • Improving transport safety
• • •
• •
• •
Urban planning Regulations for planning Provision of efficient infrastructure for intermodal transport and rail transport Supporting measures for rail / intermodal transport. Support of innovative research projects
• • •
Supporting measures for intermodal transport Energy efficient public vehicle park and vehicle technologies Support of innovative research projects Support of use of private sidings Optimisation of building material transport
The canton and city followed a strategy to reach more sustainable freight transport. On the local and regional level measurement plans for freight transport were carried out including infrastructure, technical, economical and legal measures. An important part of the local and regional freight transport policy is to support rail and intermodal freight transport. Therefore, the authorities of Zurich as well as a number of European projects are trying to achieve a higher share of intermodal transport (e.g. IDIOMA and SPIN). Compared to passenger transport freight transport is strongly determined by the market. Nevertheless, in freight transport, political framework conditions are necessary because private initiatives alone do no lead inevitably to more sustainable freight transport since the economic dimension compared to the social and ecological dimensions have a much higher weight.
INNOVATIVE INTERMODAL CONCEPTS AND CITY DISTRIBUTION Within Switzerland several intermodal concepts and technologies for short and medium distances and small and medium intermodal terminals have been developed. Some of these concepts and technologies include urban distribution. Within the IDIOMA project such solutions have been demonstrated and evaluated in the Zurich area. An integrated concept for intermodal transport and urban distribution has been developed by the Swiss Federal Railway.
Small container solution (IDIOMA Project) Within the IDIOMA project (IDIOMA, 2001; Ruesch and Glucker, 2001) the use of a small container system for consumer goods within the transport chains of the PISTOR company (between Rothenburg and Zurich, 60 km) was tested and evaluated in 2001 (Figure 9). The main objectives were to reach less-than-truck-loads in urban distribution for intermodal transport and to reduce the specific emissions per consignment. The demonstration included as main innovations small containers (for 4 pallets), a low floor city lorry, a swap body frame for 4 small containers, a horizontal transhipment equipment and telematics applications including tracking and tracing techniques. The technical evaluation showed that the technical feasibility of the system elements (city lorry, transhipment equipment and swap body frame) functioned generally but were not ready for the market yet. This equipment has to be improved as well as the telematics application equipment.
Urban rail and intermodal freight strategies in the Zurich area
373
Figure 9 Elements of the small container solution The evaluation of the economical effects showed for the individual situation of PISTOR and the today's framework conditions that the small container solution leads to higher costs than in the zero state case by road. The costs per net ton are for the road case (with small containers) about 17% to 25% higher and for the intermodal case (with small containers) 28% to 35% higher (Figure 10). Considering the low market prices for road transport and the market prices for rail transport the differences are in reality even higher. In several scenarios also the implementation of the Heavy Vehicles Fee (HVF) was included but we found that the influence was modest because of the relative low mileage in the main haul and especially in the distribution leg. The cost savings are already compensated by the higher distribution costs of the city lorry with a small container (high vehicle and personnel costs) and additional costs for transhipment have to be paid.
Figure 10 Costs per net ton for different alternatives and scenarios (Ruesch/Gliicker, 2001) If better framework conditions as longer distances, higher delivery point densities or cheaper transhipment costs can be created, the intermodal case (with small containers) would become competitive compared to road transport. The evaluation of the environmental effects showed, that the small container solution used in pure road transport chains compared to conventional road transport leads to a decrease of road mileage of about 25% and a decrease of 20% to 30% in emissions and energy consumption. That means that the external costs (emissions and accidents) are reduced by 17%. If the small container solution is used in intermodal transport chains there is a decrease of road mileage of about 65%, a decrease of 25% to 55% concerning NOx, HC, CO2 and CO2, and a decrease in energy consumption of 39%. In this case, the external costs are reduced by about 30%. Considering the results of the demonstration and the evaluation and improving the efficiency of the system we see applications for the small container solutions for city delivery (historic centres), rural regions delivery (with small container depots), delivery to shops or restaurants which need additional storage, party/events delivery services using boxes with cooling units 3 About 1 Eurocent per tkm (2003) and about 1. 5 to 2 Eurocent per tkm (2005).
374 Logistics systems for sustainable cities (which can be used as a stationary storage), night delivery services and also special applications in combination with automated underground distribution systems in cities or in development areas and also in combination with automated stocking facilities.
ACTS (Roll-on/RoII-off-System, IDIOMA) The intermodal technology based on ISO-containers and swap bodies have a very high requirements concerning terminal infrastructure (IDIOMA, 2001; Ruesch and Gliicker, 2001). Therefore, the transhipment costs in terminals with small and medium quantities are high and the pre- and end-haulage by road is usually long (low density of terminals). The ACTS System (horizontal transhipment technology with roll-off containers, Figure 11 has lower requirements for the transhipment points and is easy to used for truck-drivers (low capital and operation costs). The ACTS-System has been used in recent years in Switzerland especially for bulk goods (e.g. building materials, waste and agricultural products) over short and medium distances, also in urban areas. The system reached a remarkable market share but is currently limited to bulk goods. The main objectives of this sub-project within IDIOMA were to evaluate the experiences with ACTS equipment and also to assess its potential for further applications. Within waste transport chains the system has been used since 1997 in the canton of Thurgau over short and medium distances of about 30 to 70 km. Seven vehicles with ACTS containers are in operation.
Figure 11 ACTS System The following effects were observed: (a) Reduction of vehicle requirement (from 17 to 7 vehicles, not only caused by ACTS, also an effect of tour optimisation) (b) Share of intermodal transport of 60% in tons (close to the burning plants road transport is more suitable, before 100% road transport) (c) Reduction of costs between 5 and 10% (d) Reduction of mileage: 600'000 km a year (e) High practicability and reliability User acceptance (Zweckverband KVA Weinfelden, general public and transport companies) is very high regarding transport efficiency, competitiveness, environmental aspects, working conditions and transport safety. The system has been implemented successfully. ACTS was also used (in the year 2000) in a pilot project within consumer goods transport chains (food and non food) of MIGROS between Neuendorf and Crissier (approximately 160 km). For the first time in Switzerland a loading platform has been used on a roll on/roll off vehicle. For the end-haulage a gas-driven truck was used with low emissions and low noise levels. At the beginning of 2001 MIGROS decided to stop the pilot project because of the uncertainties about the future of intermodal systems and the low return load. But MIGROS is
Urban rail and intermodal freight strategies in the Zurich area
375
still interested in realising intermodal transport and will make a decision on the intermodal system to be used in a later stage. Within IDIOMA five different transport chains and different scenarios (with/without Heavy Vehicles Fee) in order to compare road and intermodal transport costs were investigated (Figure 12).
Figure 12 Transport cases for price comparisons The findings were: (a) Below 100 km intermodal transport using ACTS is only competitive compared to road transport under special framework conditions (e.g. waste logistics and high volumes on block trains) (b) Between 100 and 200 km intermodal transport using ACTS can be competitive compared to road transport due to the productivity gains of the weight limit for road transport (from 28 to 401) and the length of the pre- and end-haulage (c) Over 200 km intermodal transport using ACTS is usually independent of the Heavy Vehicles Fee and cheaper than road transport. But the HVF improves its competitiveness. Also the environmental impacts were analysed in regard of the five different cases (different distances and kinds of goods) and weight limits (2000/28t, 2001/34t and 2005/40t). Depending on the distance class the energy use of intermodal transport, using ACTS saves up to 50% compared with road transport. The reduction of social costs (emissions and accidents) is estimated between 30% and 50%. Overall, the reduction in mileage, emissions and social costs by intermodal transport are significant. Furthermore an analysis of the potential for ACTS within Swiss internal traffic was undertaken. It showed a potential to be shifted from road to intermodal transport with ACTS of about 3 to 5 Million tons a year for 2001 and within the time horizon 2015 a potential of about 10 Million tons due to an increase of freight transport and an improvement of the intermodal system. The system can also be used in urban areas (e.g. waste transport). There is strong competition with other vehicle based intermodal transhipment systems such as the Mobiler or the Cargo Domino system.
Cargo domino system The cargo unit of the Swiss Federal Railways (SBB Cargo) invented in 2002 under the product name Cargo Domino a new intermodal transport system based on vehicle related horizontal
376
Logistics systems for sustainable cities
transhipment equipment (Figure 13). Vertical transhipment by crane or reach stacker is also possible. This system was especially developed for consumer goods, raw and bulk materials.
Figure 13 Cargo Domino System It is now used in the inland consumer goods transport chains of 5 large wholesalers on distances between 80 and 300 km. SBB Cargo is responsible for the organisation of the whole transport chain. In January 2003 were about 65 loading units per day. In summer 2003 around 140 loading units are expected. Horizontal transhipment from rail to road is possible at railway goods stations and the infrastructure requirements are low. So existing infrastructure can be used. Today in Switzerland 11 transhipment points are in operation with Cargo Domino. At these points attractive closing and preparation times are provided. Until 2008 the number of transhipment points will be increased to 23. Thus a very dense transhipment network will be available which leads to short pre- and end-haulage distances. Operationally, the transport with Cargo Domino is integrated in a single wagon based express railway network with one central shunting yard in the centre of Switzerland. There are day and night connections available. The system is competitive to road transport from about 80 km distance depending on the individual transport chain characteristics and framework conditions. Thus also a modal shift from road to intermodal transport on short and medium distances is possible. A positive influence has the Heavy Vehicles Fee (approximately 0.5 Euro/km). The system could also be used for European inland transport.
THE ROLE OF RAIL IN URBAN PLANNING The study of freight transport within the conurbation of Zurich (Ruesch et al. 2003) besides dealing with demand analysis also dealt with rail services concepts and the identification of infrastructure needs. The background of the study was the increasing pressure on the railway infrastructure (urban goods stations, private sidings and shunting yards) within the urban area of Zurich because of the city development. One important question was also if the existing goods station Zurich Cargo has to be replaced and if the area will be needed for other purposes.
Urban rail and intermodal freight strategies in the Zurich area 377 The results of this investigation were needed for the local and regional transport plan as well as for the development plan of the Swiss Federal Railways SBB. Depending on the traffic segment the following freight service options are generally suitable:
Traffic segment
Table 2 Rail freight services options (Ruesch et al. 2003) Inter modal Transport Rail Transport
Internal traffic (within • Shuttle/direct trains from private siding to • Intermodal shuttle/direct trains between urban private siding on connections with very the region of Zurich, goods stations and very big singular traffic high volumes per day (bulk transport) max distance 50 km) generators (bulk transport) • Vehicle based horizontal transhipment equipment • Shuttle/direct trains from private siding to • Intermodal shuttle/direct trains from terminal Inbound/Outbound private siding on connections with very to terminal (conventional combined transport) traffic (Swiss national, high volumes per day 50 to 250 km) • Single wagon and wagon groups in express • Single wagon and wagon groups in express freight trains (day and night connection) freight trains (day and night connection) • Single wagon and wagon groups in normal • Single wagon and wagon groups in normal freight trains freight trains (for delivery of private sidings) • Liner trains (in the future) • Shuttle/direct trains from private siding to • Intermodal shuttle/direct trains from seaports Inbound/Outbound and important regions to terminal private siding on connections with very traffic (international, (conventional combined transport) high volumes per day 150 to 2500 km) • Single wagon and wagon groups in normal • Single wagon and wagon groups in normal freight trains freight trains (for delivery on private sidings) • Express trains (in the future) • Express trains (in the future) • Liner trains (in the future)
For internal traffic within the Zurich region only shuttle or direct trains are suitable because conventional single wagon traffic costs are too high costs and require too long lead times over short distances. For national and international transport also single wagon and wagon groups traffic in express and normal freight networks can play an important role. Liner trains and international express freight trains maybe an option in the future. From these operational options we can derive for the freight railway infrastructure the relevant functions (Table 3). The infrastructure for urban goods stations and private sidings should be very flexible relating to the possible functions. The conventional loading and unloading should be possible but also the transhipment of intermodal loading units. Also private sidings could be used as intermodal transhipment points. Private sidings can also be opened for other companies in the neighbourhood, so a more efficient use of the infrastructure is possible. For the most important locations a combination with a city logistics platform can also be suitable. Table 3 Railway infrastructure functions (Ruesch et al. 2003) Train formation
Shunting Yard Terminal Urban goods station Private sidings
Transhipment (crane, reach stacker)
Transhipment (vehicle based, horizontal, incl. small containers)
Loading/Unloading railway wagons
Additional functions (warehousing, commissioning, transport bundling)
X (X) (X)
X
X X
X
X
(X)
X
(X)
For urban goods stations after considering the demand analysis and the existing locations for the Zurich region the following main findings were determined:
378
Logistics systems for sustainable cities
(a) A concentration of the railway traffic on less urban goods stations is suitable (reduction of rail operation and shunting costs and better use of infrastructure). The locations should be easily accessible by rail and road and should be close to existing and potential clients. (b) Nevertheless quite a dense network of urban goods stations is necessary to reduce the preand end-haulage distances (c) Very important is a connection to a freight express network (consumer goods require short leading times) (d) The infrastructure for urban goods stations must have a high flexibility relating to different functions (loading/unloading, transhipment of intermodal loading units and automated storage facilities) (e) A combination with a city logistics platform can increase the volumes and the attractiveness of an urban goods station (f) A replacement of the main urban goods station Zurich Cargo in the city centre is necessary to cover the demand and to keep future options for city delivery open (delivery from the centre to the suburbs) For private sidings after considering the demand analysis and existing locations for the Zurich region the following main findings were determined: (a) Decreasing need for private sidings for production plants (shift of industrial production, decreasing company sizes and decreasing consignment sizes) (b) Ongoing need for private sidings for distribution platforms of logistics service providers or shippers (c) The infrastructure of private sidings must have a high degree of flexibility relating to different functions (loading/unloading, transhipment of intermodal loading units and automated storage facilities) (d) For industrial areas and logistics parks joint private sidings with access for third parties should be realised. Based on the results of the study the required locations for urban goods stations and private sidings have to be secured in the regional and local transport plan. These plans are being revised now by taking into consideration the results of the mentioned projects. As a supporting measure the authorities can rely on the existing law for co-financing improvements for existing infrastructure or also new locations. Other suitable supporting measures are access regulations and road pricing approaches which will increase the use of rail and intermodal transport.
CONCLUSIONS The concepts and projects in the Zurich area show that despite the change in freight demand, rail and intermodal transport can play an important role in urban distribution. Important preconditions are, quite dense network of public goods stations which allows the transhipment with vehicle-based transhipment technologies. Therefore the pre- and end-haulage distances can be kept low as well as the costs. Also suitable is the combination with city logistics platforms. Another important issue are the supporting measures such as access regulations and road pricing solutions which provides demand management and a modal shift. Integration between intermodal transport and urban distribution is possible. Central distribution structures support the use of pre-commissioned loading units. There is a potential for mode shift. Regional and urban freight strategies and concepts should therefore always consider railway and intermodal transport.
Urban rail and intermodal freight strategies in the Zurich area 379 The railway and intermodal infrastructure required has to be identified and preserved considering new production concepts and technologies. When railway infrastructure is used for other purposes, usually it cannot be rebuilt at the same place. Also the realisation of new railway or intermodal infrastructure can be useful. Thus the necessary existing and planned infrastructure has to be secured in regional and local transport plans. Rail and intermodal freight can contribute to a reduction of road freight transport and environmental burdens in urban areas. Since rail and intermodal transport are more sustainable than pure road transport, authorities should support the infrastructure and use of rail and intermodal transport. To achieve a breakthrough the framework conditions play an important role, especially the effective charging of transport (including external costs), open access to railway infrastructure, the interoperability of the intermodal systems (not too many!), agreements for building sites and businesses producing high goods volumes.
REFERENCES Amt fur Verkehr Kanton Zurich (2001). Verkehrspolitische Ziele und Grundsatze fur die Gesamtverkehrskonzeption des Kantons Zurich. Zurich, December 2001. BESTUFS (2001). Workshop on the role of rail within urban distribution. Presentations. (available on www.bestufs.net) Glucker, C. and Ruesch, M. (2001). BESTUFS: Best Urban Freight Solutions. City Inquiry ,,European Survey on transport and delivery of goods in urban areas". February 2001. (available on www.bestufs.net) IDIOMA (2001). Innovative Distribution with intermodal freight operation in metropolitan areas. Best Practice Handbook. 2001. (www.idioma.gr) Ruesch, M. et al. (2003). Study about freight transport within the conurbation of Zurich. Study on behalf of the Canton of Zurich, City of Zurich and SBB. Zurich, Final report, 14th July 2003. Ruesch, M. and Benito M. (2002). Todays and future freight transport chains. Analysis and need for standardisation. 2002. Ruesch, M. and Eugster, J. (2001). Strategic development of freight transport in the canton of Zurich. Study on behalf of the Canton of Zurich. Zurich, Final Report, 8.10.2001. Ruesch, M. and Glucker, C. (2001). IDIOMA: Innovative Distribution with Intermodal Freight Operation in Metropolitan Areas. Validation and Evaluation for the Validation Site Zurich. Project within the 4th Framework Programme of the EU. Internal Report IR 9.4. Zurich, Final Report, 31th August 2001. Ruesch, M. and Ruggli, P. (1998). Feasibility of measures for optimisation of urban freight transport. Effects of a ,,City suitable" vehicle pare. Study within COST Action 321: Urban goods transport. Zurich, 1998. Stadtplanungsamt Zurich (1995). Ziele fur den Guterverkehr in der Stadt Zurich.
This page is intentionally left blank
27
ON-LINE RETAILING IN FRANCE CURRENT AND FUTURE EFFECTS ON CITY LOGISTICS Daniele Patier, Research Engineer, Laboratoire d'Economie des Transports, France Louis Alligier, PhD. Student, Laboratoire d'Economie des Transports, France
ABSTRACT Business-to-consumer e-commerce results in deliveries to consumers. Before evaluating the impact on urban networks, it is necessary to describe and understand the underlying logistics. Certain sectors, e.g. food products, require specific solutions. Distribution systems are presented as an alternative to private shopping trips. However, the costs of order handling (order-picking, packaging) and home deliveries are still high and not always covered by the price paid by the consumer. On the whole, the profitability of on-line retailing is uncertain and the number of clients limited.
INTRODUCTION In light of estimates on the growth of e-commerce foreseeing a significant increase in the flow of vehicles to deliver goods to consumers, the goal was to determine the impact of these deliveries on city logistics. This presentation is based on a study carried out by the authors in 2002 on the logistics of various e-commerce sectors. The study was financed by the French Ministry of Transportation. The first observation concerned the small amount of data available, the lack of reliable data (published figures varied by a factor of one to five, depending on the source) and the lack of ratios used for monitoring purposes. The newness of the phenomenon provides a partial explanation for this situation.
382
Logistics systems for sustainable cities
The study consisted of two parts, first, the formulation of a typology describing the various logistics organisations set up and second, on the basis of the resulting data, the formulation of scenarios on possible future developments in e-commerce. The second phase is still in progress. It started with a "Delphi" study based on discussions with experts in view of identifying the key factors for the future of e-commerce. The next step, using the structural analysis method (Michel Godet.1999), was to determine how the various factors interact and to develop hypotheses on the future of this distribution channel. Consequently, the results presented here correspond to the first part of the study only. This part consisted of continuous monitoring of the sites and interviews of the actors involved in the chain, from the e-order to delivery distributors. Given the great diversity of situations, it was necessary to interview all the large scale distribution's responsible and their logisticians or carriers, as well as their logistics suppliers. We visited their platforms to understand their organisation. On-line consumers and experts in home deliveries were also included in the list of interviews. The retail sectors whose operations were analysed in this study are those with the highest purchase frequencies and requiring a specific logistics system, e.g. food products. The logistics cannot be handled by traditional operators, such as the postal service, for reasons that will be discussed later. Delivery of certain products includes accessory services, such as installation and set up (computer and hi-fi equipment, household appliances). These services increase the time vehicles are parked in the street. Little data is available to provide an indication on the propensity of the French population to make purchases on the Internet. In 2002 in France, Internet users represented 37% of the total population (61% in Denmark). Among the total population, 7% made purchases on the Internet (11% in Germany) and among internet users, 19% made purchases on the internet (54% in the Netherlands). The main products sold were books, music and videos, i.e. products similar to those sold in other countries. The logistics systems in the selected sectors were described as the product of an "interaction" between a number of players, namely the consumer, the web merchant, the logistician, the haulier, the infomediary. This analysis made it possible to highlight the strategies of each player and to understand the various motivations. From that could be derived the initial elements of information on the foreseeable effects of those strategies on the flows of urban delivery vehicles.
TYPOLOGY OF ON-LINE SALES Among the great variety of products offered on the Internet, a typology may be formulated according to the organisational complexity of the logistics systems. Though it is easy to offer services, whether for sale or not, the complexity of delivery increases for material goods, depending on the type and on the various recipients. If the analysis is limited to urban zones, which concentrate 80% of the total population in developed countries, the effects may cause major changes in the flows of personal vehicles and delivery vehicles. The question is to determine whether the development of e-commerce will disturb the functioning of city logistics.
On-line retailing in France: Current and future effects on city logistics
383
It is possible to distinguish between transactions using the typology presented below. (i) Purchases of immaterial services. Information, discussions, viewing timetables, displaying times, tourism information, cultural events, etc. No material goods must be delivered. Only the method changes, whereby the internet replaces the telephone. There is no effect on the flow of vehicles. (ii) Purchases without delivery. Train, plane or show tickets, which are picked up where they will be used. This type of transaction is, a priori, without any effect on the flow of vehicles if payment is also made on line. If not, a trip may be necessary to deliver the tickets. The on-line purchase may avoid a trip by the consumer to make the purchase. (iii) Purchases of general mail-order products. These are the traditional mail-order products, such as clothing, games, records, DVDs, interior decorating etc. The internet replaces a paper catalogue and a trip by the consumer to the store is avoided. Deliveries are generally made via the postal service or by transportation companies that centralise and organise the flow of products. This operating mode does not create any additional flows of vehicles. If the web merchant organises deliveries on its own, there is a risk of greater flows of delivery vehicles. (iv) Purchases of less frequent products requiring delivery and installation. Televisions, computers, furniture, household appliances etc. It is possible to imagine the following scenario: - 1 trip by the consumer to see the product in a store; - 1 trip by a carrier or by the web-merchant vehicle; - 1 trip to install and set up the product by an expert. The scenario is identical to that for a traditional purchase, but it is possible to avoid a trip if the purchase is made directly via the web site, without a prior visit to a store, (v) Sale and delivery of groceries to consumers. General goods, hygiene products, dry, fresh and frozen food. This type of transaction is the most complex because it brings into play the formidable problems concerning the cost of the last kilometre, the integration of the logistics systems of very different sectors, delivery times and just-intime stock management. The success of this type of transaction depends entirely on the quality of the logistics and on the strategic decisions made by all the players concerned with supply chain management. The following table compares the share of e-commerce turnover with the number of sites offering each type of product: Table 1 On-line retail sales in France during 2001. % of e-commerce retailers % of e-commerce sites turnover offering this product Food 1% 23% Beverages 3% 16% Clothes 28% 15% Records 15% 38% Computer equipment 2% 13% Sports 4% 13% Software 1% 10% TV, Video, Hi-fi 7% 10% Games, toys 1% 10% Books 0% 9% Flowers 2% 8% Services 9% 8% Type of product
384 Logistics systems for sustainable cities Beauty products Motor vehicles Others
1% 0% 2%
8% 1% 18%
Source: Insee, April 2001.
TYPOLOGY OF THE PLAYERS INVOLVED Interaction takes place between five, well-identified players; the on-line consumer, the web merchant, the supplier, the logistics company and the carrier, and also the infomediary. The diagram below presents each player, their main actions and the flow of information and physical goods.
Figurel Interaction between players To avoid dealing with excessive detail in this presentation, we will not discuss the spatial and institutional aspects, even though they were, of course, taken into account in the study itself. For example, the on-line consumer can order from home or from the office. He can also choose the place of delivery and the delivery time. His decisions are also influenced by the town planning and regulation background. In the course of the study, the following players were identified:
On-line retailing in France: Current and future effects on city logistics
385
Table 2 Players E-consumer Web-merchant: pure players (1) and click and mortars (2) Supplier Logistician and hauliers Home delivery specialists Infomediary
Our sample 20 e-consumers (1) top achat; Mr good deal (2) exhaustive hypermarket sites (5)+ FNAC, IKEA, RungisMIN, Picard... Venditelli, La Poste, UPS, ... Ikaris, Stars'service, TEC Team on line , Cross log
E- Consumer This is the first link in the chain. The behaviour of consumers, their reactions and expectations influence the strategies of the other players (web merchants, logisticians, carriers). There are two subgroups: • Young, unmarried persons (male or female), urban, aged around 30, with higher education; • Two-income households, high revenues, children, belonging to the higher socioprofessional categories for whom time is a decisive factor. Far from being isolated, on-line consumers live in a city, have a car and relatively easy access to nearby stores. They shopped in major chain stores or in shopping centres before starting to make purchases on the internet.
Web Merchant This is a highly diversified group whose behaviour depends on past experience. Starting in 1995, the number of sites grew sharply. There has been a strikingly high number of failures due to an underestimation of the difficulties raised by the logistics and the low level of profitability due to the lack of a true strategy. We find start-ups, pure players, mail-order companies, large volume distribution, small traders, independent producers, hypermarket, major chain stores Mail-order companies • these are the experts in mail orders and home deliveries made throughout the entire country. They have the clear advantage of perfectly mastering supplychain management, from the initial order on through to individual delivery. They already have a mechanised order-processing platform, a computerised system for returned products, etc. They are also the pioneers in delivering to "drop-off points" where the client can come to pick up the order (the major mail-order companies such as La Redoute, Les 3 Suisses, CAMIF, etc.). These companies simply put their catalogue on the internet, with all shipping taking place via the standard channels, i.e. the postal service or "integrators". There was no change in the delivery system. Hypermarkets: five groups cover the market. All of their internet sites opened between 1998 and 2002. In 1998, Telemarket (Galeries Lafayette group) already had 16 years of experience in handling out-of-store orders and home deliveries and was thus on an equal footing with the mail-order companies. The C.mescourses company (Casino group) was launched in 1999 and
386
Logistics systems for sustainable cities
closed in 2002. Ooshop (Carrefour group) and Houra (Cora group) opened in 2000. Auchan was the last to open a site in 2001 and profited from the errors of the others. Retail traders and other small stores: these stores took advantage of the internet to expand the territory covered and to present their products rather than to sell them. Their capacity to sell is limited by the difficulties involved in the logistics. Consequently, sale volumes are low. They are trying to develop an efficient mode of operation. The genera] policy is to make every effort to deliver to homes immediately, within two hours after check-out in the store. There have been efforts to create a common site for a dozen merchants in a given city (often in the city centre) who thus pool their delivery capacities. Major chain stores (FNAC, IKEA, etc.): these stores simply went on line. That changed nothing in their upstream logistics systems. However, a complex organisation had to be set up to ensure compatibility between the existing deliveries from stores to homes (with installation of equipment) and home deliveries of products ordered via the internet from dedicated platforms. Supplier The supplier can be a central purchasing department, a production platform, a large volume distribution platform or a dedicated platform...The supplier receives the information immediately when the consumer orders on line. He has to organise the restock for the logistics companies and to deliver them just in time. He must be able to restock the logistician or the haulier platform according to the volume of orders, in order for efficient, real time Supply chain Management. Logistics companies and carriers This highly competitive market is clearly segmented. We find integrators, including the postal services with their subsidiaries and the express services, independent carriers and home deliveries specialists (for the last kilometre); Integrators and express services : five major companies share the world market, namely DPWN (DHL and Euro Express), TNT, Consignia , UPS and FEDEX. Their advantage lies in the capacity to propose a complete set of services (storage, packaging, packing, tracking, aftersales service, returns, etc.) and world-wide coverage (networks). However, they accept only single packages with weight (less than 30 kg) and size (less than one square metre) limitations. They are active in both B2B and B2C. The market is currently undergoing significant integration due to a complex system of agreements. Among the integrators, the postal services (via their subsidiaries) play an important role. In response to globalisation, they have entered into partnerships with the large integrators (the French Poste/FEDEX, Deutsche Post/TNT, etc.). They are the leaders in home deliveries. The subsidiaries of the French Poste, (Chronopost, Colipost, Esipost) are involved in B2C. None of these integrators offer delivery of fresh or frozen food. Independent transport company : these companies hesitate to enter the home-delivery market (too expensive) and have difficulties handling fresh food. Messenger services, active players in
On-line retailing in France: Current and future effects on city logistics
387
deliveries to urban zones, also propose a number of other services, including warehousing, E.D.I., tracing, booking of orders and packaging/packing. Home delivery companies (for the last kilometre): new players on the market, following the failure of traditional carriers in the B2C sector, these companies already offered delivery immediately after check-out in grocery stores. Accustomed to transporting fresh and frozen food products, they have adapted perfectly to the requirements of e-commerce. There is currently a vast number of companies being launched to provide home-delivery services and they are attempting to penetrate the e-commerce market. In France today, the leading home-delivery company has 1000 vehicles every day in the streets of Paris. Infomediary Though invisible for the consumer, the infomediary acts as a supervisor and is the only participant to have an overview of the entire transaction. He is the "invisible actor", but he plays an important part in the informational chain. He builds the software which permits the immediate transmission of all the characteristics necessary for the orders which converge on the web-merchant site towards all the others actors. He triggers all the organisation of the successive deliveries between suppliers, logisticians, carriers, consumers and the warehousing provisions, the delivery time slot.
THE POSSIBLE EFFECTS OF E-COMMERCE ON CITY LOGISTICS Effects are produced by a number of factors, including: • the number of trips for products transport; • total distances travelled; • the number of vehicles and their size; • the road occupancy of both moving and parked vehicles; • how vehicles park. These effects depend on: (i) the behaviour of on-line consumers in terms of travel habits, particularly for purchases. Their lifestyle, access to stores and to the necessary technology (computer and internet link) play an important role; (ii) the logistics systems of the major web merchants, i.e. platform sites (distance from the delivery zone), the number of players and their co-ordination, the number of transfers via platforms and the grouping of tasks, the decision to fill orders (picking) in a store or on a special platform, the proposed deadlines, the number of products (and different types of products) offered and the management mode (in-house or subcontracted); (iii) the organisation of carriers and logistics companies. The success of a site depends entirely on its logistics.
388
Logistics systems for sustainable cities
PLAYER STRATEGIES In this new market, interaction between the players is intense. It is characterised by very fast action, reaction, adaptation, market swings and changing demands in terms of service quality. This new market was made possible through the use of new information and communication technologies (NICT). Each player has a strategy based on very precise criteria that set off real-time reactions on the part of the other players. By observing the sequence of action presented in figure 1, it is possible to compare their expectations as well as their strengths and weaknesses, and consequently the strategies implemented. The figure 2 below presents the relation between the main players with their strengths and weaknesses.
Figure 2 Strengths and weakness of e-commerce players The on-line consumer calls the shots Home deliveries have met with great success over the past few years. In the year 2000, they represented 5% of food purchases for French households, i.e. 380 M€. Some 20% of households (including 12% at least once per month) use the service, where two thirds are for goods from stores and one third from cyber-markets. The average order represents just under 110 euros. The typical consumer profile is a two-income family with high revenues, children, middle-aged and the means to "out-source" household tasks. Before shopping on the Internet, the family shopped in large stores.
On-line retailing in France: Current and future effects on city logistics
389
A particular characteristic of France is the relatively low percentage of internet purchases compared to other countries due to the existence of the Minitel, a "mini internet" used exclusively in France. Even today, 51.7% of out-of-store orders transit via the mail, 35% the telephone, 9.5% the Minitel and only 3.8% via the internet. Retail sales on the internet represent only 0.05% of total retail sales and only one seventh of those made via the Minitel. The delivery of food products to homes is an urban phenomenon given the difficulties for consumers, i.e. the distances to large stores, the time lost in traffic and the difficulty of parking in front of the home to unload the goods. Also, this type of offer is possible only if the demand is sufficiently concentrated. In the year 2002, it represented only 2% of out-of-store purchases, but it is the sector most likely to encounter strong growth. It is also the sector most likely to push web merchants to modify their strategies and to create disturbances in city logistics. The study revealed the expectations and hesitations of on-line consumers. Consumers want a wide range of products, quality and prices at least equivalent to those in stores, short delivery times (two hours to next day) and within specific time slots, punctuality (±15 minutes), transaction traceability, delivery information (interaction between the call centre, the delivery person and the client) and secure payment (currently the most limiting factor, with the possibilities of legal recourse). Obstacles include lack of access by households to computers, the percentage of computers connected to the internet (preferably via high-speed lines) and the relatively low level of computer proficiency. In the year 2000, 12% of French households were connected and the estimate for 2003 is 27%. In November 2000, 28% of households declared they were ready to make purchases via the internet. Increases in equipment levels and the propagation of new purchase habits should result in the internet representing 5 to 10% of purchases over the next ten years. The on-line consumers questioned tend to group their purchases of convenience goods, which increases the price of the average order, but they are not particularly loyal. They tend to go after the best offer and demand is very volatile. The greatest advantage expressed by consumers is the time saved. They estimate that on average, depending on the distance from their home to the large stores, they save 90 minutes for each order via the internet compared to a store purchase. In terms of distance, they are of the opinion that they save 10 to 20 kilometres in their car for each order via the internet. The time saved is devoted primarily to their families, to activities at home or, in some cases, to free-time activities.
Web merchants Start-ups, small producers and small merchants do not have the means to establish sophisticated logistics systems. They must either make deliveries themselves, on request, or out-source to the postal service or a traditional transporter (generally a messenger service). For supermarkets and major chain stores, the goal in creating a site is primarily to be present on an innovative market and to be a leader on that market. There are many difficulties :
390 Logistics systems for sustainable cities (i) Create a logistics system corresponding to the needs and purchase habits of consumers. (ii) Optimise geographic coverage. For supermarkets, the only location without any risk is Paris. There are two possible strategies, either a gradual spread outwards, or immediate national/international coverage with the possibility of pulling back if results are not up to expectations. (iii) Select the least expensive logistics systems, (iv) Optimise the products sold on line, given that large numbers of products complicate the logistics system and increase the risk of client dissatisfaction, (v) Target clients and make them come back, anticipate demand and adapt to new demand (niche markets), (vi) Improve communication to reassure clients.
How do they operate? In light of the many difficulties listed above, the strategies implemented by major chain stores were fairly similar, whereas supermarkets used very different strategies. There is clearly great uncertainty and continuous changes in direction. These fluctuations have obviously resulted in adaptations on the part of the other players, notably the logistics companies. Most of the major chain stores have their own means of transportation to supply the stores in Paris and for home deliveries. They have set up different logistics systems for e-commerce. The e-commerce platform is supplied like any other store. They offer delivery under 24 hours via a subcontractor who responded to a call for tenders and is considered a true partner. Ecommerce deliveries go through the postal service and for home deliveries, drop-off points are increasingly common. For deliveries abroad, Chronopost (express postal service) would seem to be the most suitable solution. For supermarkets, the first strategic decision is which clientele to target and to offer the number and types of products likely to satisfy that group. Currently however, the potential clientele is very limited (see the characteristics above). The groups that attempted to cover the entire country have since pulled back. Only the very largest French cities are now covered. Certain internet sites serve only the Paris region. The most difficult points are the logistics and transportation. Filling the order remains the most expensive step and the most difficult to organise. Here as well, there are a number of very different strategies. (i) Filling the order. Sites hesitate between picking in-store, on the central platform or on a dedicated platform. The advantage of picking in the store is that the upstream logistics are not at all modified, but store operations and inventory management are severely disturbed. The advantage of the central platform lies in the fact that it is perfectly integrated in the supply-chain management, however, it is often too far from the city for deliveries. That explains the trend toward dedicated e-commerce platforms. Construction of these platforms is extremely costly and the return on investment is not only uncertain, but very long term under the best of conditions. (ii) The location of platforms is critical. They must be located as close as possible to city centres to ensure a high level of service and reduce the distances covered by vehicles, their size, fuel consumption and the time spent on transport. In terms of size, a platform must be capable of filling one to two thousand orders a day in 2003, with capacity for five thousand if necessary. They occupy vast amounts of space, from 2 500 up to 18 000 square metres
On-line retailing in France: Current and future effects on city logistics
391
and require a large workforce (200 to 250 employees, generally young, for repetitive and tiring work). A balance must be found between a location as close as possible to the city centre, with its advantages and disadvantages (very expensive real estate, high turnover of labour), and a location farther out offering less expensive land and better labour conditions, but with the disadvantages mentioned above. Most dedicated platforms are located within ten to thirty kilometres from Paris. All the larger companies are now studying the question from the sustainable-development angle. The most desirable platforms are those abandoned by the national train company (SNCF) because they are located in the city centre and may, if all the players can co-operate, serve for multi-modal transportation with trains supplying the platforms, particularly for dry food products. (iii) Operating mode. A number of decisions (mechanised and automated processes, use of new information and communication technology, order consolidation, packing) impact directly on the downstream logistics systems. Mechanised processes require a major investment and total control over data exchanges. The quality of the computer systems and the software employed determine not only service quality and reliability, but also the relation with the last link in the chain, i.e. delivery. (iv) All web merchants started out by filling orders and making deliveries on their own, then some of them slowly out-sourced these tasks. There are now a number of different organisations. • All operations remain in-house. Filling of orders is carried out by company personnel who also make deliveries with company vehicles. • Some operations remain in-house. Orders are filled on the central platform and products are then transported in company vehicles to a dedicated platform. Deliveries are handled by an outside transporter. The dedicated platform is the equivalent of a store for the upstream logistics system. Transportation downstream is optimised by organising regular delivery routes that avoid numerous trips by consumers to stores. • Partnership with a logistics company. Orders are filled by company employees on the platform of the logistics/transportation company. This type of platform is the equivalent of a company store (this solution was used by C.mescourses on the platform of the transportation company Venditelli, which also made the deliveries). • The second type of partnership is with a logistics company. The web merchant receives on its dedicated platform a number of logistician executives from the logistics company to organise order filling and transportation (selection of the subcontractors). This is the case for Telemarket.
Logistics/Transportation Companies These players operate in a context of heated competition and must deal not only with subcontractors providing low-quality service, but also the insufficient number of truck drivers. However, when companies are large enough and sufficiently well organised, the rewards can be high. This is the last player in the supply chain and the one to make or break the reputation of the web merchant in the mind of the customer. There is no room for error.
Numerous constraints A reduction in the lead time between the order and delivery has become a key competitive factor. The company capable of delivering faster and better will win. (i) Fast execution, same day or next day, two hours after check-out in grocery stores.
392 Logistics systems for sustainable cities (ii) Control over computer systems. (iii) Control over logistics. (iv) Own and manage a fleet of vehicles that can be adapted to the needs of clients. In particular, select vehicles suited to urban conditions, with two or three temperature zones for fresh and frozen foods. Fleets are increasingly leased, thus providing flexibility for shifts in demand, (v) Control over the last kilometre and even the last metre (interphones, door codes, blocked entries, absent clients), (vi) Control over delivery dates, (vii) Punctuality (appointments for delivery), (viii) Responsiveness, adaptability, (ix) Anticipation of needs (vehicles and drivers), (x) Personalised delivery services. Different and more or less effective organisations Own Account delivery : These systems continue to exist in parallel with out-sourcing. Auchan has its own delivery system. Telemarket has gradually given up its system, considered too complex to meet the many demands of the market. Some 30% of deliveries are still made with in-house systems to limit transportation costs and to better evaluate the offers of logistics companies (comparison of both the services rendered and the cost). Third party with an exclusive agreement :This implies a true partnership. The major advantage is the possibility of setting up unique solutions satisfying both partners. On the other hand, it is a very dangerous situation because if one partner does not perform, the other fails as well. But the challenge can be stimulating. The transporter is authorised to operate on the home-delivery market for grocery stores, but may not work with another web merchant. The web merchant is also restricted to the single partner (this is the case for C.mescourses with Venditelli and Stars' Service with Ooshop). This type of partnership produces the most effective results. Certain carriers have over 1 000 vehicles in Paris every day, exclusively for home deliveries, of which 10% are e-commerce deliveries. Routes are organised using special software and vehicles are monitored in real time using geographical-tracking software. Third party with several suppliers: Telemarket and Houra decided to go with several suppliers. Telemarket was accustomed to this organisation due to its experience in taking orders via the telephone and the Minitel. However it is difficult to manage a larger number of suppliers, particularly in terms of hour volumes and the equitable distribution of cargo. These problems are made all the more difficult due to major variations in daily volumes. The second danger concerns the lack of control over further subcontracting, thus bringing into play other suppliers whose responsiveness, adaptability and quality of service are difficult to ensure. This explains the current reorganisation with Telemarket reducing the number of suppliers from 24 to 3 in 2003. This type of organisation may, however, be necessary for a web merchant covering the entire country. An example is Houra which, with 60 000 products, often delivers very different types of products in a single order. By selecting messenger services well positioned on transportation lines, it is possible to use their platforms to split packages in the provincial cities. Traditional carriers'- Finally, for most general products, the postal service or any other integrator remain the best solution due to their organisation in networks, worldwide coverage and, for the postal service, its specialisation in home deliveries. The postal service is already heavily involved in mail-order activities with order-filling platforms positioned near major
On-line retailing in France: Current and future effects on city logistics
393
sources of activity (La Redoute, L'Oreal, etc.). Now, it is used for products that previously transit via mail-order channels, given that 80% of e-commerce orders were previously made by telephone or Minitel, and the remainder consists of new clients. All postal services have set up portals, but that has not resulted in more packages being sent.
FORMULATION OF SCENARIOS The results derived by formulating scenarios are approximate, given the small amount of information available. The only goal can be to establish hypotheses on the effects of ecommerce on city logistics. The newness of the phenomenon deprives the study of any quantitative data and it is clear that the decisions made by the players are by no means final. More solid results will be available only after further monitoring of players and the methods employed. At this point in the study, it is possible to say that: (i) the strong growth of e-commerce is based on immaterial goods, information and various services; (ii) time will pass before e-commerce significantly modifies consumer habits; (iii) for some time already, a majority of the high-volume products sold on the internet have been handled by well-organised networks with the capacity to absorb a significant increase in volume; (iv) in terms of logistics, e-commerce has not modified existing networks. Large companies such as La Redoute process postal and internet orders in the same manner and still deliver as they always have. Mail-order companies remain the major players; (v) the postal service is capable of gradually accepting higher volumes of packages. It is possible to assess the effectiveness of different scenarios. Scenario 1. Delivery of food products by a supermarket This scenario is based on interviews with e-commerce players in France in 2002. One logistics/transportation company Some 180 vehicles make 15 deliveries per round through Paris, covering an average distance of 30 kilometres. Per day, they make a total of 2 700 deliveries and cover 5 400 km. On-line consumers A total of 54 000 vehicle-kilometres would have been travelled if the 2 700 consumers had made the purchases themselves, covering 20 km round trip (average distance in Paris). In addition, for over half the distance, the vehicles would have carried no goods. Result Home deliveries avoided 48 600 vehicle-kilometres. Even if only half the consumers used a car, the savings still amount to 21 600 vehicle-kilometres per day. Scenario 2. Deliveries by major chain stores with services Stores have their own fleet of vehicles to supply their stores in Paris and make home deliveries limited to Paris. In terms of logistics, the e-commerce platform is considered simply another store and does not affect the general organisation of the logistics system. Small packages are shipped via the postal service, using Colissimo for home deliveries and Chronopost for deliveries abroad. Deliveries to drop-off points are increasingly common. Result Deliveries are made via traditional circuits, no additional vehicles are on the road.
394
Logistics systems for sustainable cities
Scenario 3. Deliveries by grocery stores after check-out This is by far the least effective scenario. Consumer The consumer collects and pays for the goods. The check-out person consults the delivery schedule, proposes a delivery time and calls the delivery company. Company specialised in home deliveries The call centre in the delivery company takes the call and sends an employee to the store to package the goods. The employee travels with an empty vehicle from the call centre or home to the store. The company employs poorly qualified personnel, has no plans to optimise operations, but provides personalised service to specific demands. The vehicle goes directly to the client's home and comes back empty. Result Numerous trips for a single client, vehicle empty half the time. On the basis of our study, it would appear that: (i) consumers are in the most favourable position; (ii) consequently, it is unlikely that they would accept to pay more for the service rendered (order filling and delivery); (iii) the level of profitability required to improve the quality of logistics systems and maintain operating costs (e.g. mechanised order filling) depends on achieving scale economies. Given that demand is still low, the return on investment is not only uncertain, but very much long-term under the best of conditions; (iv) among the on-line purchase-decision criteria, the time factor is today the key element (Chronopost International [2001]). The consumer makes purchases on the internet primarily to save time. If one accepts the commonly stated idea that the value of a person's time is proportional to the person's revenue, the e-commerce clientele is limited to the most affluent part of the population; (v) under current conditions and for the sectors covered by the study, generalised use of the internet to make purchases is not likely.
CONCLUSION It would appear to be necessary to gain a wider perspective on the observed phenomenon, hi terms of a service including home delivery, on-line B2C sales do not constitute a new market. The previous existence of mail-order operations and the very low impact of e-commerce on the purchasing habits of consumers are very instructive when speculating on the future of business on the internet. That being said, is a shift in purchase-decision criteria not already underway? Previously, the popularity of mail-order was based primarily on the originality of the products offered. However today, one of the advantages of on-line sales is to considerably reduce the time required to make a purchase and, in addition, to make purchases possible at any time. Will these advantages become important for a wider segment of the population and a permanent factor in decision making? Furthermore, it may be assumed that there is a relation between the mobility of people and their propensity to want home deliveries. The typical consumer for the e-commerce sectors studied is a working person with a high-mobility profile and a tight schedule. If for environmental reasons, mobility in urban areas is limited (e.g. restricted access to city centres for individual vehicles with a corresponding increase in the use of public transportation), other, wider segments of the population may well turn to on-line purchases and home deliveries.
On-line retailing in France: Current and future effects on city logistics
395
REFERENCES Patier, D. (2002). La logistique dans la ville. Celse. Brousseau, E. (2002). Impact des TIC sur les modeles logistiques du commerce electronique : nouveaux metier, nouvelles formes d 'intermediation. FORUM pour le PREDIT. Latourte, C . (2002). Les consequences urbaines du e-commerce -I'exemple des cybermarches franciliens. Institut Francais d'Urbanisme. Universite Paris VIII. Browne, M. (2001). Commerce electronique et transport urbain. OECD/ECMT. Les effets du commerce electronique sur les transports. Session 4. Transport et distribution locale. Nemoto, T, Visser, J., Yoshimoto, R. (2001). Impacts of information and communication technology on Urban Logistics System. OECD/ECMT The impact of e-commerce on Transport Session 4 : transport and local distribution. CREDOC (2001). Livraison a domicile, le cas de Paris et de la petite couronne. Browne, M. (2000). e-commerce freight distribution and the truck industry. ACEA. Rallet, A. Consequences du developpement de nouvelles formes de relation au client final sur I 'organisation de la chaine logistique. IREPP pour le PREDIT.
This page is intentionally left blank
28 DISPELLING THE E-COMMERCE AND URBAN TRANSPORT ENVIRONMENTAL DOOMSDAY FORECASTS: A COUNTER INTUITIVE AUSTRALIAN CASE STUDY - THE POSTAL TRANSPORT NETWORK RESTRUCTURE, 1995 TO 2000 Kim Hassall, Associate Professor, Department of Civil and Environmental Engineering The University of Melbourne; Director Urban Trucking Strategy, Australian Trucking Association, Australia
ABSTRACT Since the McKinsey Company first forecast, in 1994, that home shopping via the internet, would change our lives forever, there has been a continual concern that this demand would generate both more Business to Business (B2B) road transport trips, and a quantum leap in the Business to Consumer (B2C), householder delivery trips. This was also additional to the growth expected in most major capital cities over the next fifteen to twenty years. Although there is much discussion as to when the internet began to generate, or most likely substitute existing business into either smaller B2B pulses to customers, or beginning the household e-Commerce deliveries, by 1998 the internet was generally accepted as having had an impact. In Australia over the half decade, 1996 to 2001, demand was growing, however there were some astonishing and counter intuitive fleet developments emerging within the largest transport network operator in the Australia, namely, the domestic Postal Corporation. In the face of increased workloads, the fleet size fell in three transport segments: long distance prime mover operation, urban rigid operations, and within the light commercial vehicle sector. Fleet vehicle numbers fell 14%, kilometres travelled by 2% and fuel consumption by 7% in this fiveyear period.
398 Logistics systems for sustainable cities Overall a large incumbent network transport and delivery operator may well be able to draw upon economies of network density to, in fact, become far more efficient than most commentators would believe. The re-configuration of the fleet vehicle types and new fleet productivity measures drove these changes. The changes relied on changes to several fleet policies and changes in the construction of the serviced operations network. Hit words: Urban e-Commerce, B2C householder delivery, e-Commerce and the environment, environmental transport case studies, urban logistics, urban transport environment, urban deliveries, householder deliveries Australia.
BACKGROUND The following case study reflects the complete opposite to what most forecasters would have predicted across the five year growth scenario in an urban transport task during the eCommerce boom. It reflects the true operational changes on the road transport task of Australia's major B2C e-Commerce processing and delivery agent, across the period 1996 to 2001. In 1996 B2C e-Commerce was beginning to happen although many commentators do not actually begin measurements of B2C until 1998. Irrespective of the "agreed" commencement point of e-Commerce, the data sets in Tables 1 and 2 span the actual commencement date of B2C network operations within Australia, and the impact has been captured in the task presented in these to sets of data. The first observations were raised in Hassall, May 2001 at the OECD, Paris e-Transport summit. Transport consolidation on this observed network was actually beginning to show very much improved outcomes against the 1996 base period. What actually occurred between 1996 and 2001 within this urban fleet was a significant growth in task capacity. This added capacity in fact drove a decline in the number of commercial vehicles required to do the task. The net number of commercial vehicles actually fell by — 14.2%. The net kilometres dropped slightly and fuel consumption declined by 7.4% for these commercial vehicle groups. Table 1 Vehicle and Fuel Usage: 1996 Vehicle type Light Commercial Vehicle Medium 2-axIe Rigid Heavy 2-axle Rigid Medium 3-axIe Rigid Local Articulated Linehaul Articulated Total
Gross Vehicle Mass (tonnes) 2.0-2.5 11.9-13.0 15.0 22.5 39.0 42.5
Number of Vehicles 1602 428 175 87 41 48 2381
Fuel Consumption Rate (I/I 00 km) 16.1 21.9 27.3 31.3 47.1 38.3 ..
Total Annual Travel (million km) 40.43 13.60 11.74 8.19 4.15 12.90 91.01
Total Litres (litres) 6 491 768 2 978 400 3 205 020 2 563 470 1 954 650 4 940 317 22 133 625
However, what drove the changes within the fleet mix, and hence a very desirable business efficiency and environmental outcome? The outcome was based on four underlying and concurrent processes.
Dispelling the e-commerce and urban transport environmental doomsday forecasts
Vehicle type
Table 2 Vehicle and Fuel Usage, 2001 Number Fuel Gross Vehicle of Consumption Mass Vehicles Rate (tonnes)
Light Commercial Vehicle Medium 2-axle Rigid Heavy 2-axle Rigid Medium 3-axle Rigid Local Articulated Linehaul Articulated Total
(1/lOOkm)
2.0-3.49 11.9-13.0 15.0-16.0 22.5 - 23.0 39.0-42.5 42.5 - 63.0
1384
..
2043
222 172 190 27 48
12.3 21.7 27.2 29.5 39.0 44.2
Total Annual Travel (million km) 39.50 9.30 10.06 16.84 2.57 11.10 89.37
399
Total Litres (litres) 4 858 500 2 018 100 2 736 320 4 ^ 7 SIX) 1 002 300 4 907 968 20 490 988
OPERATIONAL TRANSPORT FLEET BUSINESS SENSIBILITY The fact that two heavy rigid trucks could replace three lighter rigid vehicles. As can be seen in Table 3, the lighter and medium rigid vehicles 11.0 tonnes GVM to 16.0 tonnes GVM were declining in numbers to the larger three axle rigid trucks operating up to a 23.0 tonne GVM capacity. It was also true that two long wheel base vans could replace four short wheel base vans, which had half the capacity of the long wheel base variety of light commercial vehicle (LCV). Light commercial vehicles dropped in overall terms by -13%.
THE SWITCH TO A DIESEL LCV The fuel consumption rate for the light commercial vehicle class began to drop significantly as one-tonne GVM petrol LCVs were replaced with diesel vans. The average consumption of the 1996 petrol vans were 16.1 Htres/lOOkms. By 2001 the replacement of the petrol LCV by its larger diesel equivalent was only 12.3 litres/1 OOkms. This was a positive reflection of the efficiency of the diesel engine for urban work despite the perceived bad reputation of light urban diesel vehicles. Table 3 Chan;;e in Vehicle Numbers, Use and Fuel Consumption, 1996 to 2001 Change (per cent) Gross Vehicle Vehicle type Fuel Total Mass Number of Total Consumption Annual Vehicles Litres Rate Travel Light Commercial 2.0 - 3.49 -13.6 -25.2 -23.4 -2.3 Vehicle Medium 2-axle Rigid 11.9-13.0 -48.1 -0.9 -31.6 -32.2 Heavy 2-axle Rigid 15.0-16.0 -1.7 -14.6 -0.4 -14.3 Medium 3-axle Rigid 22.5 - 23.0 118.4 -5.8 105.6 93.8 Local Articulated 39.0-42.5 -34.1 -17.2 -38.1 -48.7 0.0 Linehaul Articulated 42.5 - 63.0 -0.7 15.4 -13.9 Total -14.2 -7.4 -1.8
400
Logistics systems for sustainable cities
THE IMPLEMENTATION OF LIMITED FLEET OPTIMISATION SOFTWARE TOOLS The introduction of certain types of fleet optimisation software occurred from 1994 onwards, but the application of such tools was only at the margin. It did however provide valuable productivity benefits in the problem areas that the models were focused. The resource savings, through the use of optimisation software, ranged from 13% to 30% on many and varied localised problems.
NETWORK RESTRUCTURING Lastly, and possibly the most significant factor in the reduction of the net kilometres and fuel usage, was in the redesign of the postal transport network in the two largest cities. In brief in 1996, in both major cities, six medium processing centres all serviced each other with their transport network. This resulted in an N2 number of transport links. By 2001 the six processing centres had been consolidated to three larger processing facilities each with a localised cluster of small consolidation and transfer centres which serviced each major processing facility. The case study is insightful as it runs counter intuitive to the forecasts of all the urban transport and environmental transport pundits. Even in Australia whilst forecasters were expecting a doubling of some segments of transport activity (NTS, 2001d) this major network was reflecting a contraction in network entropy, fleet activity and kilometre activity. The original multi processing facility network had large numbers of direct customers and post offices connected directly to a mail processing facility, with each processing facility essentially having its own catchment area. Such a structure brought with it a high degree of inefficiency with regard to the utilisation of fleet resources.
Dispelling the e-commerce and urban transport environmental doomsday forecasts
401
Figure 1 The Abstract view of the Original Multi Processing facility network Figure 2 reflects the network restructuring outcomes for the post 2000 mail network. The process involved the creation of a new type of node that performed customer and post office collection, some degree of product streaming and consolidated despatch to a mega facility. These new centres, some 15 such facilities in the two major capital cities, in actuality are called hubs, although they are realistically mini pre processing and consolidation centres. The six node processing mail centre network reflected in Figure 1, was in fact consolidated into a two megacentre network pictured in Figure 2. In fact the mega-centre network, with its star "mini streaming and consolidation centres" was a device to avoid the growing impact of vehicle congestion at particular critical time windows. It also provided some time savings at the major processing centre. The customer and retail centre consolidation at these mini processing centres demanded a requirement for larger rigid trucks. This drove the requirement for the smaller rigid trucks to decline.
402 Logistics systems for sustainable cities
Figure 2 The Abstract view of the new mega and mini centre processing network The main products handled by the new centres are B2C letters and parcels, although small to medium B2B business is also attracted. The mega centre reconfiguration was to take advantage of new sorting technologies although the hub and mega-centre network allowed a significant degree of pre processing to occur at the new larger feeder hubs. In essence, the new network was designed for growth and this certainly enabled the 'newer" B2C business, where it was identifiable, to be accommodated.
Dispelling the e-commerce and urban transport environmental doomsdayforecasts
403
Table 4 Advantages and Disadvantages of a single major centre network Advantages of a mega centre processing facility: Some degree of pre-sort and consolidation of product can be achieved before delivery to the mega centre can occur, The mini consolidation centres help diffuse the impacts of city congestion, The vehicle tied requirements reflect feucr but larger trucks. Centralisation of the workforce may achieve an absolute optimised level of staffing. Ids-advantages of a mega centre processing facility The network of mini centres is capital intensive The associated processing risk of having one mega centre: e.g., industrial disputation, lightning strike, electricity blackouts or terrorism may have a devastating impact on the entire postal network. Changes in local area regulation may have a severe impact on the operation of a single processing facility. The one centre may not provide a quick response to important B2B clients at the other geographical extremity of the capital centre
CONSIDERATIONS FOR A "MULTIPLE" OR A "LIMITED" PROCESSING CENTRE NETWORK The philosophy of a single or multiple facility processing network has various operating benefits but also comes with some potential negative considerations. These are presented in Table 4. What a large national, or even a public common carrier, can achieve is highly dependent on risk, capital outlay and planning for either, one or two large centres, as opposed to multiple processing centres; in this case six centres. A further consideration is the placement of the supporting transport centre, or depot, that supports the mega processing facility. Ideally the large transport centre housing at least the heavier truck fleet would be co-located adjacent to this processing facility. It would also have access to bulk fuel and a highly utilised truck and light commercial vehicle maintenance facility. There is considerable weight in the argument to have the Light Commercial Vehicles (LCVs) out stationed near smaller consolidation centres and nearer to customers. These parking arrangements also reduce the kilometre requirements for daily first pickups from a centralised mega centre.
SUMMARY OBSERVATIONS This case study raises very interesting issues for the urban transport policy designers, especially for future research on the impacts of common public carriers. It reflects the benefits of consolidation as well as ensuring benefits from a "better" designed urban distribution network. For a network task that increased significantly over the five year period there was a consolidation achieved that reduced vehicle numbers significantly in some segments of the fleet, namely LCVs and light trucks, 11.9 tonnes to 13.0 tonnes GVM. Network kilometres travelled decreased marginally by minus 1.8%, and fuel consumption by minus 7.4%. These results certainly present a refreshing scenario of possible achievements through integrated transport network planning. With slight changes in rigid truck dimensions and weight carrying capability further productivity gains, in excess of 11 % are achievable through the adoption of
404
Logistics systems for sustainable cities
rigid truck Performance Based Standards (PBS). (Hassall, 2003). This adds yet another dimension to achieving the possibility of planning for a sustainable urban transport environment up to the year 2020.
REFERENCES Hassall, K. P. (2001). Trends and Hindrances in e-Logistics: An Australian Perspective, The Impact ofe-Commerce on Transport, OECD/ECMT Seminar, OECD June 2002, Paris. Hassall, K. P. (2003). Achievable Rigid Truck Productivity Gains through Performance Based Standards, International Seminar on Performance Based Standards, NRTC Melbourne. Hassall, K. P. (2001). The Evolution of B2C, Supply Chain Review, December 2001, PSA, Brisbane. Hassall, K. P. (2002). The e-Logistics challenge for the Post Office, Proceedings 7th International Symposium on Logistics and Operations Strategy, Melbourne July 2002. Hassall, K. P. (2002). What e-Logistics environment will a Small and Medium Enterprise expect by 2006?, Freight Transport Industry Action Agenda, National Office of the Information Economy and Department of Transport and Regional Services, Canberra. National Transport Secretariat, (2001a). e-Business Impact on Transport Systems, Working Papers 1 to 8, NTS, Brisbane. National Transport Secretariat, (2001b). Introducing The Transport Impacts of E-Business Project, Working Paper 1 of the E-Business and Transport Project (ref 252), NTS, Brisbane. National Transport Secretariat, (2001c). E-Business Trends, Working Paper 2 of the EBusiness and Transport Project (ref 253), NTS, Brisbane. National Transport Secretariat, (200Id). Transport Impacts And E-Business, Working Paper 3 of the E-Business and Transport Project (ref 254), NTS, Brisbane. National Transport Secretariat, (2001e). Insights From Stakeholders/Experts, Working Paper 4 of the E-Business and Transport Project (ref 255), NTS, Brisbane. National Transport Secretariat, (2001f). Data And Data Source Issues In Assessing The Transport Impacts Of E-Business, Working Paper 5 of the E-Business and Transport Project (ref 256), NTS, Brisbane. National Transport Secretariat, (2001g). Regional Impacts of E-Business in Australia, Working Paper 6 of the E-Business and Transport Project (ref 257), NTS, Brisbane. National Transport Secretariat, (2001h). Ranking & Rating The Transport Impacts of EBusiness, Working Paper 7 of the E-Business and Transport Project (ref 258), NTS, Brisbane. National Transport Secretariat, (2001i). Opportunities For Transport Related Productivity Gains, Working Paper 8 of the E-Business and Transport Project (ref 259), NTS, Brisbane. OECD (2003). Performance Based Standards for Road Transport, OECD, Roads and Transport Research Program, Paris.
29
E-COMMERCE AND END DELIVERY ISSUES
Marielle Stumm, Inrets, Arcueil, France Daniel Bollo, Inrets, Arcueil, France
ABSTRACT
E-commerce businesses have been undergoing rapid development for the last five years in the United States and for the past two years in Europe. This sustained growth illustrates the existence of a demand for this type of service, particularly among the youth. Beyond the startup phase, e-commerce companies are continuing to generate significant losses, which point to organisational defects, the most serious being logistic support to this business. Analysis of the e-commerce issue is delicate, given the haziness of the activity's perimeter. E-commerce startups offer services similar to traditional mail-order, and consumer retailing is not clearly stating its objectives in creating its own e-commerce sites. Logistics is not an organisational technique that is adapted to the rapid and unpredictable changes that e-commerce is experiencing today. Logistics related problems in e-commerce vary according to the type of commercial activity involved, but they are often considerable and sometimes result from the precipitation with which these activities were set up. Home delivery service expanded to grocery purchases, as offered by a number of major retailers, is an even more sensitive issue. In the current state of the market and the organisations implemented, the cost of delivery exceeds 8 Euros and often leaves customers dissatisfied with the often approximate conformity of the order. The situation can only improve with the organisation of truly adapted delivery systems, which will most likely require commercial pooling actions.
406
Logistics systems for sustainable cities
THE EXTENSION OF THE SERVICE APPEARS TO BE POOLY MASTERED Home delivery is not an especially innovative service. There is an entire retail sector based on this service and most convenience grocery stores deliver to their customers on request, particularly the elderly. But retailing continues to evolve to adapt to a demand which, itself, is rapidly changing. Consumer retailing in particular is feeling the expectation for home delivery from its clientele, to the extent that it is progressively extending this service throughout retail outlets, with a variety of solutions. At the same time, the rapid development of e-commerce is increasing the number of transactions involving direct customer delivery. Some transportation carriers have specialised in this area, locally in France, while in the United States, there exist national operators (HomeDirect) or large operators which have opened specialised services (e.g. Fedex and UPS). Solving the delivery problem may seem trivial, while it is responsible for serious difficulties among sales staff in a number of cases. Today, it is the customer who ensures this shopping pick-up service. Transferring this service to a third party would generate relatively substantial additional costs. Worse, consumer retailing is heavily organised to receive customer visits. Considerable expenditure is devoted to decoration, special events, etc. Reversing the roles is not easy and will require time. One might think that groceries in the largest sense {consumer retailing) is a special case and not very representative of the products which are effectively delivered to the home. This is rather true today. But if we place ourselves within a perspective of development, it is clear that, by controlling a good portion of retailing, this sector will probably generate the largest portion of this service to private consumers. Here, we should note that nearly all of the major retailers have set up an e-commerce site. They are making considerable efforts to position themselves in this promising sector. It seems that adaptation to online services is truly difficult. There is currently not a single ecommerce site that is making money. On the contrary, losses are high and they are progressing as quickly as the companies' revenues. Over the last three years, the world leader in consumer retailing, WalMart, has remodelled its offer several times and is now being forced to close its site for several weeks or more, in order to drastically reorganise it. The difficulties encountered are not solely related to delivery problems, but these remain a significant liability. Although there exist relatively efficient organisations, and although costs naturally vary depending upon the product delivered, it seems that the figure of 12 Euros per delivery in urban areas should be considered as the lowest in statistical terms. A symptomatic evolution can be observed in the United States. Free delivery and fast delivery were the rule among most e-commerce sites three years ago. Today, customers are often charged for delivery, and its price varies according to enticement grids that take the speed of service into account. But, will these retailers be able to keep their customers, when these customers have become used to a free service? Whatever the difficulties, this service is regularly progressing to the extent that it is generating concern over urban traffic and its corollary, pollution. But the answer to these questions is not very clear. The traffic generated by buyers going to a store is as great, or greater, that that
E-commerce and end delivery issues
407
generated by delivery trucks, all the more so as the family model includes fewer and fewer members.
OVERALL DIAGRAM It is very difficult to analyse the current delivery situation. We are probably in a transition period in which the different interested parties are simultaneously testing various approaches. This diversity only reflects the large number of parameters, whose progressive mastery should enable the establishment of standard solutions. A conventional analysis of the solutions proposed using statistics, is not pertinent here. These solutions are too recent and, given the poor financial results, their promoters are understandably reticent to detail the problems they involve. Insofar as it is highly probable that this sector will continue to grow, it could be of interest to analyse the factors present and to determine possible avenues for at least approximately solving the problem. Ideally, the solutions proposed should simultaneously satisfy the different parties involved in the issue. The retailer would like to offer an attractive service at a reasonable price, the carrier would like to operate under reasonable conditions, the customer would like their desires taken into consideration and public authorities should take care to keep this from causing traffic problems. Delivery to the customer is not the final link in the logistics chain. Upstream supply is at least as important and must be paired with distribution in order to form an efficient supply chain. For reasons of clarity, we will treat only final delivery here. This latter issue is often presented as the "last mile" or even the "last yard". The ingredients of the problem are few, but the combinations are quite considerable. Figure 1 below illustrates the problem
Figure 1 Basic delivery diagram Illustrating the different parties involved is not the sole purpose of this representation. The problem is obviously not only topological. Home delivery of newspapers is viable because the product to be distributed is unique and the area over which the clientele is distributed is
408
Logistics systems for sustainable cities
relatively dense. Moreover, it also reduces the cost of unsold papers, which are very high in the press industry. We repeat, there is no single solution any more than there is a single commercial branch. Both sectors are in fact highly correlated, yet distinct. In the case of the press, there are two channels: home delivery is often handled directly by the publisher, but delivery to points of sale is usually handled by specialised distributors. We may also observe that the different players may have multiple roles: (a) (b) (c) (d)
A customer may order from several retailers for the same period A distributor may handle logistics for a colleague company Certain warehouses may have several retailers in common An order may require retrieving supplies from several warehouses
Finally, delivery conditions in downtown areas are varied and may be successively satisfied via: (a) (b) (c) (d)
Direct remittal to the buyer Deposit in a special box Leaving goods at a warehouse Remittal to a third party: building superintendent, concierge, etc.
CUSTOMER EXPECTATIONS For a clearer definition of the delivery parameters, it is useful to assess customer expectations in terms of retailing. There are two observations: (a) consumer habits are constantly changing, as can be easily observed through comparisons with 20 years ago, and (b) there is a wide diversity of customers and even a variety of customer attitudes towards a product. This does not facilitate analysis, but we can note that today, retailing is dominated by very large groups operating in large suburban shopping malls. This illustrates the fact that retailing is seeking a size effect, like many other business sectors, in order to increase turnover, particularly in logistics, and appeal, through the diversity of its offer. In concrete terms, ecommerce is confronted with a powerful, efficient economic system, that the consumer appreciates. Furthermore, this same consumer in the year 2000, who has relatively high buying power on the average, is also looking for sophisticated, branded products. We can observe this in the growth of sales in trendy products and heavy appliances (PCs, laptops, games, stereos, etc.). hi this case, the effect of mass marketing is less marked and leaves room for products with high added value for which consumers accept to pay more for an occasional need. E-commerce is more comfortable in these sectors, all the more so as the product sought is not necessarily available locally.
E-commerce and end delivery issues 409 The differentiation in consumer attitudes can also be found in Internet usage. The under-thirty population often has the reflex of first checking the internet for the product they are looking for, if only to forge an opinion on the current offer, in terms of existing product features, availability and prices. They may then be tempted to make on the online purchase directly. Furthermore, we may observe that e-commerce is strictly related to Internet availability in households. In Sweden, the world's highest level of Internet access finds its corollary in the maximum volume of sales via this channel. There are also the usual factors to explain sales: (a) revenues (b) available time (c) mobility A couple of young executives with children dispose of income that enables them to accept an additional expense related to e-commerce and will appreciate the time saved with home deliveries. At retirement, this same couple could consider shopping as a form of entertainment, as long as they are still mobile (Table 1). We must also observe that the notion of available time is rather variable. To overcome the problem of working couples, shopping malls have considerably extended their business hours and opening days by remaining open later or by opening on certain Sundays and holidays. We could add that these malls work hard to attract the public by multiplying special events, shows, decorative themes, etc. Not only are there no longer many obstacles in terms of hours, but going out to shop on Sundays could be a form of entertainment. Mobility limitations would seem an advantage for e-commerce. But only in appearance, as convenience retailers have a firm grip on this market segment. Local grocers make home deliveries to the elderly in the neighbourhood, with the added advantage of quality contact that only long practice has enabled them to build. Table 1 Appeal of E-commerce Appeal of E-commerce Little availability Large availability + Normal income ++ ++ +++ High income + ++++ ++++ (Two cases per square: normal / limited mobility) This analysis should be taken a step further, but that is not our point in this paper. Moreover, it is of interest to observe more generally that the remote retailing clientele is neither particularly typed or even loyal. Industry surveys show that all socio-professional categories and all regions use remote retailing services and only for occasional purchases. On the other hand, it appears that e-commerce operates with a clientele essentially composed of young households.
410
Logistics systems for sustainable cities
LIMITATIONS OF THE PRODUCTS SOLD Delivery problems naturally differ according to the type of product distributed. That is why the logistics sector is largely segmented by product: beverages, meat, chilled products, etc. This is not problematic when it comes to delivering large quantities to super or hypermarkets. The end-customer is potentially an eclectic consumer, but who buys small quantities, which singularly aggravates the delivery problem. The primary factor is the purchase price as compared to weight and volume. It is clear that costly products do not pose delivery problems, as delivery costs are only a small portion of the price paid. Heavy appliances (e.g. television sets, stereo systems and computers) sell well via remote retailing. The delivery portion represents only a few hundredths of a price that is moreover highly variable from one retailer (or brand) to another. This reasoning is also relatively applicable to lightweight products with small volumes such as clothing, which make up a significant portion of traditional mail-order revenues. Their small volume also facilitates storage with a third party, such as a building superintendent, if the consumer is absent. We could continue with examples such as books, CDs, etc., to arrive at the observation that the good cases of home delivery were already being exploited by the existing companies, well before the irruption of e-commerce. By expanding the scope of application of mail-order systems, this pioneering e-business sector can only encounter difficulties, while such difficulties have been largely ignored by their developers. Generalist sites, in particular, making home deliveries of family groceries have to take four categories of products into account, for which the imperatives differ, and which therefore cannot be combined very easily: (a) Dry goods (b) Produce (c) Chilled products (d) Frozen products Specially adapted trucks are required for home transportation, featuring three compartments, including two that are temperature controlled. This imperative also makes it impossible to prepare orders at the same time at the warehouse. Sometimes, these restrictions result in the products being stocked in different warehouses. This is often the case for produce. In addition, logistics are always related to a products' packaging. Products are delivered to the home by unit, but leave the production plant on palettes, which in turn are composed of cartons containing several units of the product. While logistics are well organised for handling standardised storage units like cartons and palettes, it has traditionally left final handling of the products to store managers. That is the whole difference between the forklift driver and the handler (Figure 2).
Figure 2 Units and handling systems diagram
E-commerce and end delivery issues
411
As the final assembly of the products to form an order requires extensive handling, specialists are seeking a standard transfer unit. This could again be cartons, but the absolutely heterogeneous aspect of the contents of a delivery is not conducive to this form of packaging. Another solution also tested consists of using specially adapted solid storage units, containers, on which consumers pay a deposit. Some are even insulated to keep certain products cold. The disadvantage remains volume, which is often an obstacle in densely populated urban areas.
DELIVERY CONDITIONS Although home delivery is a traditional technique, the pressure of e-commerce has driven intensive and extensive experimentation of a large number of solutions. An examination of these solutions, whether recent or long-standing, innovative or well proven, shows that there are too many parameters to enable the emergence of a universal solution, or even the establishment of general principles, which only experience of a certain length of time will enable us to validate. In addition to conditions related to the products themselves, the organisation of a delivery depends upon packaging factors that have their own imperatives (Table 2). The places in which the packages and products are grouped and ungrouped.
Place Warehouses traditional Warehouses home delivery Retail traditional Retail home delivery Depots
Table 2 Organisation of delivery and packaging Entry (objects / Exit (objects / vehicles) vehicles) palettes palettes trucks trucks cartons palettes vans trucks products palettes cars trucks cartons palettes vans trucks cartons cartons cars vans
Objects handled palettes cartons cartons products cartons products cartons products cartons
It is important to note here that it is difficult for the two functions to coexist — deliveries to stores and to private consumers from the same warehouse, due to the differences in size and volume of the objects handled. It is difficult, and even dangerous, to have fork-lifts and picking carts circulating in the same area. Two possible strategies are currently being explored. Either we adapt the existing warehouses to accommodate both functions, which appears to be the European direction. Or, solutions specialising in the end-customer must be set up, as can be seen through several referential examples in the United States. This last solution is relatively costly and investments may become quite substantial in the case of an automated warehouse. Experience proves that this option is very advantageous in the area of service, in terms of flexibility and rapidity, but that initial costs are extremely high (several dozens of millions of dollars).
412
Logistics systems for sustainable cities
Contrary to this, there are very few disadvantages to using stores as warehouses. There is no fundamental difference between the operations made by an intermediary and those made by ordinary customers. In this case, the picking is done early in the morning at the same time as re-shelving, and during slow hours of the day. However, it is a pure handling operation, which can nonetheless be computer assisted by establishing optimised lists of products for removal, possibly even for several customers. The cost of the operation is therefore considerable and uncompressible. The difficulty in composing delivery parcels increases rapidly with the number of product references in a catalogue. Whatever the organisation, the size of the warehouse increases proportionally with the choice left to customers and the corresponding movement within the warehouse can increase even more. We should note that the rare successes in e-commerce in strict terms of profitability lie in rare and costly product niches that are available from very succinct catalogues. However, in terms of revenues, these niches represent a significant portion of remote sales and are even more significant when compared to ordinary retailing. An extensive selection guarantees the retailer certain attractiveness, but this is inevitably linked to problems of stock costs and handling difficulties, in other words, to logistics, that are more difficult to master. The area covered by delivery from a warehouse or store is an important parameter, the denser the concentration of customers around a warehouse or store, the shorter the delivery rounds. The action perimeter is a crucial parameter. The grocer has only a few yards to cover to deliver his customers, while the e-retailer must have his delivery service cover considerable distances. In cases of sizeable dispersion, they have no other solution but to use public carriers.
DELIVERY ORGANIZATION Given the number of parameters governing home delivery, the number of possible and even existing solutions is rather high. But when we speak of deliveries, other delivery parameters must also be selected: (a) Direct remittal or not (b) Home or depot delivery (c) Delivery times set by the customers or by the distributors
Direct remittal or not It may be inconvenient to wait for a delivery, or very difficult in the case of working couples. Customers may prefer delivery to a box or lockers for storing parcels. This is often the case in the United States where the deliverer has a pass key for the garage, which acts as a personal depot. Another solution consists in purposely using a mail box that is large enough to accommodate large size parcels. This is also the case in rural areas in France for postal parcels, but there is no equivalent in urban areas. However, the characteristics of the French mail box are insufficient for voluminous parcels and suffer from security problems. Conversely, some customers are attached to direct contacts, which may be an advantage for the retailer, on the condition that the deliverer also handles commercial actions. This is often the case with existing home delivery systems for groceries. The signature of a delivery slip also
E-commerce and end delivery issues 413 solves one of the problems of remote retailing by considerably reducing the number of claims. But for working people, this postpones deliveries to late evening and weekends. Delivering essentially between 6:00 p.m. and 10:00 p.m. and Saturdays is not conducive to an economic delivery organisation. Home or depot delivery Home delivery is usually difficult when the customer is not at home during business hours. The disappearance of concierges and the generalisation of door codes inhibit delivery traffic. But once these obstacles have been solved the issue of where to leave the parcel must be addressed? In Switzerland, it is left on the entryway doormat, but is this possible in France or Italy, without it disappearing? The alternative is to leave the parcel in a depot where the customer can pick it up at their convenience. There are numerous possibilities, such as: (a) (b) (c) (d) (e) (f)
The building superintendent if they accept it A neighbourhood retailer At the workplace A real specialised depot A vendor's relay shop At the post office
None of these solutions are really very suitable. The two problems are business hours and legal responsibility for the delivery. Remote retailing uses this solution, but it is not generalised. Delivery day and times E-commerce has put the accent on the rapidity of delivery, but this does not necessarily correspond to a need. All the more so, as this rapidity comes at a high price, this will have to be paid for at some point. The cost is not only monetary, as the multiplication of express deliveries can only increase traffic problems and pollution. However, if the customer accepts delivery on the day and time determined by the delivery service, it then becomes possible to organise truly efficient delivery rounds, with considerable related energy savings. Thus, if it is agreed that weekly grocery deliveries are delivered at an appointed time, the energy savings are substantial as compared to individual travel to a shopping centre (Table 3).
Delivery Express Normal Set
Table 3 Energy cost Energy (as compared to individual travel) ++
+ -
Independently of these three parameters, significant improvement could be made by pooling solutions. If a deliverer acts on behalf of several vendors, they can shorten their rounds through the combined increase of the density of their customers. For the same reasons, a warehouse that
414
Logistics systems for sustainable cities
is shared by several distributors can provide gains in terms of deliveries. But this poses the sensitive problem of alliances which must not compete with one another. The other solution would be for the transportation industry to organise more directly for home deliveries in order to offer a range of specialised services for all e-commerce. We can mention the press as an example, which is distributed via very efficient specialised channels.
CASE STUDY: THE POST OFFICE AND E-COMMERCE Introduction The parcel and mail delivery market is changing rapidly throughout the world. Post offices are changing with deregulation. Important reforms have been passed in many countries like the Netherlands, Germany and Sweden. Many are no longer limiting their services to simple mail and small parcel deliveries. They are diversifying their range of services before competition from the integrators (e.g. Fedex, UPS and DHL). In some countries like Canada and Sweden, post offices have rapidly adapted to the electronic and digital age. In Europe, some of these are private or in the process of privatisation. They form alliances, and/or buy out, and/or invest in foreign logistics companies to offer international coverage, or to increase their services with packing or storage offers. The Dutch post office, which was privatised in 1994, bought the Australian integrator TNT in 1997 to become TPG. It is believed to have bought out twenty-four express parcel firms. It has just formed an alliance with the Japanese logistics firm, Kintestu, as part of a strategy for an incorporated network. The German post office, Deutsche Post, which since 1994 has been transformed into a joint stock company, acquired Air express international a year ago, and has the ambition of becoming the leading logistics company for 350 million people, with earnings over the last three years of 9 billion Euros. It has bought Danzas, the Swiss logistics firm, the French messenger service, Ducros and has taken holdings of 25% in DHL to acquire a global offer. The British post office, which is still a public company, but which is also preparing to become a joint stock company with public capital, has bought two German delivery services, German Parcel and Der Kurier. It has also acquired the French transportation and logistics company Courrier International. It plans to invest 2,30 billion Euros for the purchase of other foreign logistics firms to also position itself on the global offer segment. The monopoly of the Swedish postal system was repealed in 1993, but the government has remained the owner of the organisation. The Swedish postal system has nonetheless kept 95% of the Swedish postal market thanks to its technological superiority. There currently exist 90 private Swedish companies registered as postal carriers. The French post office appears to be lagging behind its competitors. The monopoly of parcel distribution has been reduced through the European directives. The change of its public establishment status is not on the agenda. It does not have the resources like the German or Dutch posts to buy out delivery and logistics companies with international coverage. For the
E-commerce and end delivery issues 415 past year, La Poste has been seeking alliances abroad, either with UPS or Fedex. Agreements have been concluded with Fedex. Before the aggressive tactics employed by its neighbours, it created a Parcel and Logistics holding, including all of its specialised subsidiaries (Chronopost, Tat Express, Publi-Trans, Eurodispatch and Soficolis) in order to maintain its third leading rank among European parcel carriers.
E-commerce services Logistics According to a study by Andersen Consulting, only 12% of the products purchased online are delivered on time in Europe. It often appears that many of these e-retailers have not carefully examined logistics and have practically forgotten that an online transaction usually results in the delivery of the product to the buyer, who usually expects it within a short period of time. European postal systems are massively invading this market with hopes of substantial profits. With their distribution networks and technical skills in managing the last mile, the postal systems are positioning themselves to face the emergence of retail sites, but for some, it is often with a handicap, international deliveries. The Dutch postal service, TPG Group, has committed itself to provide logistics support for eretailers. It is becoming a serious competitor to integrators like UPS and Fedex with its international coverage. It has negotiated agreements with the Portuguese, Greek, Spanish, Italian and French postal services, and has made alliances with Royal Mail (UK) and the Singapore postal system and it can now handle parcel shipments to 200 countries. Logistics services to e-commerce make up 16% of its revenues (40 billion). These revenues are increasing by over 20% annually. It handles stocking, packing and delivery for most of the trade sites in the Netherlands. The Swedish postal system appears to offer the most comprehensive European delivery service for e-commerce. Seventy-one percent of all products ordered are delivered within a week. The Swedish postal system has been entirely reorganised to satisfy e-commerce and offer logistics services for Scandinavia, the Baltic regions and North-western Russia. Its e-business service is covered by PostNet, PostLogistick, Post Brev and PostGirot. Its services cover small parcel shipments for businesses or private individuals, as well as 3PL (Third Party Logistics) services. Through its web site, the Swedish post office provides online tracking, cost calculation, and compiles and prints all of the documentation related to the shipment. The French postal system has extended its logistics offer to satisfy e-commerce, through its subsidiaries, Publi-Trans and ChronoPost. Publi-Trans offers future e-merchants comprehensive services for BtoB and BtoC: site hosting and design, supply, stocking, order preparation, transportation, billing and payment recovery, sales administration and returns management. This subsidiary handles 2.5 million orders for 400 customers, and 130,000 references for 52,5 million euros in revenue. ChronoPost has a specific pricing offer for e-commerce, instead of basing rates solely on volume. ChronoPost sets up an interfacing system with the e-retailer's site enabling the parcel to be tracked by customer number. Most carriers like Fedex and UPS have already been offering such services for a number of years. For France, the service appears reliable. Its agreements with Fnac, that has set its sights on becoming the French-speaking leader in online cultural product sales, is an example, but its international service appears much less efficient
416
Logistics systems for sustainable cities
despite its agreements with TNT and the Belgian, Greek, Portuguese and Italian postal services, as compared to the German postal service. The German post office, a private company recently listed on the stock market, intends to become the leader in parcel transportation and logistics in Europe, similar to UPS in the United States. Through the buyout of operators like Danzas, or its 25% holdings in DHL, the German postal system is completing its service offer to move into e-commerce by satisfying both BtoB and BtoC. It has earned over 9.2 billion euros in revenue in three years, thanks to its increased range of services and especially to its international coverage. A world-wide network that should promote the explosion of e-commerce, it claims. It will use its postal know-how for final deliveries in Germany and for financial services. The Swiss post office also offers a global e-commerce solution. The Swiss group is putting its logistics skills from mail-order to use in e-commerce, from order taking, delivery, all the way to payment. Other components of the e-logistics offer have been added to these activities: (a) A Post Office Call Centre registers telephone and written orders for the retailer (b) Modules: stock status, quality control, delivery and packing are included in warehousing (c) For payment reception, the post office account is debited or credit card payments are transmitted (d) After-sales departments handle returns and questions, and manage debtors (e) The evaluation and statistics division informs the retailer of his earnings and orders It has created the subsidiary, IPEC (Informatics Platform e-commerce), which is the sole ecommerce contact point. The Swiss post office is the only one in Europe to work for an online supermarket. This retail site, called "Le Shop" covers all of Switzerland: 3,000 references, 20% monthly growth with 70% loyal customers who can be traced over several months. The post office manages all of the logistics for this site for non-perishable goods (distribution and warehousing). Customers placing their orders before 4:00 p.m. on Day X are delivered on X + 1 by the post office. In the evening, sales figures are sent to the post office for products on stock for delivery the next day. To meet the needs of e-commerce and home buying, the British post offices offer retailers not using their services for delivery, the possibility of using the 18,000 post offices as delivery drops. Similarly, parcels returned to them due to customers not being home, are sent to the local post offices where they can be picked up the next day. Naturally, this is a paying service. The national post offices have the best logistics offer, at the least cost for e-commerce dealing in lightweight, small volume products. They have mastered the last mile through their acquired competence in mail distribution and are capable of handling the related financial flows. In 90% of cases, post offices deliver most products of this nature ordered over the Net. The post offices have the human resources required for this kind of distribution. The American post office employs 960,000 people for its national coverage, while UPS employs only 320,000 people to pick up and deliver its mail and parcels around the world. The French post office manages 310,000 workers, including 89,000 mailmen. But few postal systems are capable of offering efficient international coverage. Therefore, eretailers use integrators to cover their international needs for the speed and quality of the service these provide and for remitting the parcel directly to the interested party. But, if the customer is absent more than three times, the parcel is returned to the depot. Thus, the British postal system offers of making its offices available to act as delivery depots.
E-commerce and end delivery issues 417 The structure of the transportation networks increasingly involves the use of intermediaries for dispersing the merchandise. This trend has resulted in the extreme operating diagram known today as "Hub and Spokes": a central platform and radiating service. The integrators, for whom managing end-distribution is the most costly link, are more and more inclined to delegate this final distribution to qualified carriers, like post offices, which in turn handle grouping, ungrouping and transport via major international thoroughfares. All of the operators are aware that there is a market to be conquered insofar as traditional logistics chains have proven ill adapted to the needs of e-commerce, particularly for satisfying BtoC. Retail sites in BtoC With the arrival of the Internet and the possibilities offered by the web, post offices have proposed "cyberSales" solutions to commercial companies that have been specially designed to help them take advantage of the Internet and gain an edge over their competitors. In setting up e-business sites, these offices handle a portion of the logistics for small parcels and often financial flow management. The Swedish post office appears to be the European and even world-wide leader for its ebusiness site. It hosts one thousand e-stores and receives from 5 to 8 million visits per month for a Swedish population of 9 million. This can be explained by the fact that the Swedish population is one of the best equipped in personal computers. Fifty-two percent of all Swedes own a PC at home and among these, 48% have Internet access. The Swedish postal service delivers most online purchases, and its e-retail subsidiary, Postnet, handles payment and financial flows. It also hosts travel sites like the railroad and hotels. Although current e-retail income accounts for only a small fraction of total earnings (2.8 billion Euros), heavy growth is expected within next few years. The French post office has opened a retail web site featuring 350 shops (gourmet, cultural, home, sports, fun, fashion, luxury, high-tech, business and practical tips). It offers logistics to commercial companies that want to set up through its subsidiary Publitrans and delivery through its Colissimo and Chronopost services. France Telecom and La Poste are working in partnership to offer Web retailers comprehensive single-point service, ranging from online order taking to end-customer delivery. France Telecom's "Telecommerce" offer lets customers manage online order placing, secure payment, order tracking and delivery activation. While La Poste provides its transportation solutions, enabling delivery at special rates in 24 or 48 hours. The Telecommerce platform is secure and customers can track the status of their delivery online for most of the sites. The Swiss postal system has launched an e-commerce portal for online purchases. It offers a range of efficient comprehensive services to commercial and private customers wishing to take advantage of e-commerce. With 5 other partners, it has invested some 200 million Swiss francs in this portal. The offer includes communication (e-mail, mail-to-paper and secure mail), directories, online advertising and e-stores, as well as postal services for logistics and payment traffic. In doing this, the Swiss post office has achieved part of its strategy which consists in developing its services and putting them online in order to affirm itself on the market with offers that meet the new needs of its private and commercial customers.
418
Logistics systems for sustainable cities
The German post office has also opened a retail Web site with Cyber-shops in the field of games and entertainment, computer technology, health, career, home, etc. The service is called eVITA and receives 2 million visits per month. It features 100 permanent shops in partnership, but a total of 2,500 sites may be reached via this site. Like the Swedish postal system, the European systems are joining forces with specialised partners to install their Web sites, to create retail sites to handle part of the logistics and distribution markets for these sites.
Mail and other services With the European directives, European post offices will lose their monopoly on the distribution of mail weighing less than 100 grams, which for most represents the loss of a major portion of their revenues (between 10% and 20%). Thus, in the age of e-mail, post offices have developed a series of services enabling the dispatch and reception of mail in secure electronic form that has the same legal value as paper mail. The Swedish, French, German, Swiss and Italian post offices enable private and commercial customers to use their sites to send mail, receive bills, and to pay these bills directly via the Internet. Like the French post office's directory service, the Swedish post office, in partnership with CityMail, has created the online directory "AddressPoint". This directory includes internal corporate directories. In addition, for people who are moving, it offers online services such as real estate, construction, cleaning, etc. Post offices have offered banking and insurance services for a long time. Most European post offices give their customers (private or businesses) the possibility of managing their accounts online from their Web sites in a totally secure and confidential environment. They offer a wide range of electronic services like payment and cash management (receipts, validation and payment scheduling via BtoB or BtoC EDI). The leaders in this field in Europe are the French, Swiss and Swedish post offices.
CONCLUSION The problem of home delivery is difficult to delimit, as it includes a large number of parameters related to the customers, the products and the organisation of the delivery rounds. However, this same e-commerce is generating an increasing volume of home sales. Volume is certainly the main criteria for the implementation of specifically adapted solutions. The effect of size will enhance the creation of departments specialising in home delivery. We have observed the same phenomenon with express transportation, whose progression has resulted in the concentration of highly efficient specialised operators. E-commerce is creating major opportunities for logistics players, but it is also imposing a redefinition of their services, particularly in the complex and lucrative area of BtoC. The delivery of products purchased online to the end-customers is not a priority for the distribution giants like Fedex or UPS, for example.
E-commerce and end delivery issues
419
The postal services of the different countries, whose strengths lie in the pervasiveness of their networks, but which suffer from the cultural handicap of their rigid status and, especially, their national coverage, are attempting, through take-overs and/or interest taking and/or agreements with transportation and logistics companies to extend their field of action to satisfy the needs of e-retailers and their customers. These postal services are very competitive on the last link of the chain. They are more familiar than anyone with home delivery. Moreover, they have a large number of post offices in which to deposit a parcel when the consumer is not at home. As a conclusion, it can be assumed that electronic commerce currently has no influence on the deliveries in cities because of its small less than 1% level. However, it probably will go on growing, to reach significant levels. But the integrated management of logistics through the modern systems of information already is changing the schemes of the professional retailer. The interaction of these factors suggests that urban logistics will be modified in the long term. These structural changes are long and depend on external factors to transport, such as the rate of Internet equipment, and the familiarisation of the consumers with these tools. It is true both for consumers and for small and middle sized companies such as traders. E-commerce is most successful in Sweden where the equipment rate is higher than the rate in the United States, beyond 50 %. Concerning the familiarisation with the Net, it could be a phenomenon of generation. In that case it would be a thirty years cycle as the introduction mentioned: 1945-70 the development of supermarkets, 1970-2000 the development of hypermarkets, 2000-2030 the implementation of electronic commerce?
REFERENCES Cairns, S. (1998). Delivering the future? Centre for Transport Studies, University College, London. Kamarrainen, V. (2000). Supply Chain for e-Commerce and Home Delivery in Food Industry. Helsinki University of TechnologyKlewegt, A.J., V.S. Nori and M.W.P. Savelsbergh, (1999). The Stochastic Inventory Routing Problem with Direct Deliveries. Georgia Institute of Technology. Kuusela, S., (1998). The Postman Always Clicks Twice. Research ebusiness 2.0, Sept. 1998. Hibbard, J., (1999). Going Postal. Red Herring magazine, Sept. 1999. Nemoto, T., J. Visser and Yoshimoto, R., (2001). Impacts of Information and Communication Technology on Urban Logistics System. Punakivi,M., Saranen, J., (2000). Identifying the Success Factors in e-Grocery Home Delivery. Helsinki University of Technology. Saul, D. (2000). How are postal services in the E-word?. Action Point, June 2000. Sisodia, R. S., (1996). Threats and Opportunities in Changing Retail Environment. Capital Partners International, April 1996. Sparks, L. and A. Findlay (1999). The future of shopping. Institute for Retail StudiesWeeks, C. (1999). The importance of Logistics to E-Commerce. Bentley College. University of Texas (2000). Measuring The Internet Economy. La Tribune -Supplement Multimedia (1999). La chaine logistique, maillon vital du commerce virtuel, 22/04/1999. The Postal Record, (2000). Vol. 113, N°. 6, The postal e-volution series, June 2000.
This page is intentionally left blank
30
WEB-BASED TRANSPORT EXCHANGE SYSTEMS IN JAPAN AND ITS IMPLICATION TO TRAFFIC VOLUME
Ryuichi Yoshimoto, Senior Researcher, Systems Research and Development Institute of Japan, Japan
ABSTRACT The logistics cost is about 8 % of GDP in Japan and the transport cost is 60 % of logistics cost And also, the road freight traffic volume is 40 % of total vehicle kilometres including passenger car. Therefore, it is very important for our society to increase the efficiency of road freight transport. Carriers generally have two ways to decrease road freight traffic volume. First, is to increase loading rates of LTL vehicles by joint delivery and the second is to decrease trucks running empty through the use of a information sharing system for TL vehicles. Recently, many Websites for information sharing are established in order to decrease the empty vehicles. However, most of TL truck carriers are small companies and lack of financial and human resource for information sharing. In 1998, only 25 % of these carriers used Personal Computer for their business. In 2002, this figure increased to 80 % based mainly on the diffusion of Internet-accessible mobile phone. As the results, the possibility of information sharing between carriers is increasing. The traditional information sharing systems have been used the specific communication line and closed system during carriers. Recent systems use the Web site and quasi-open type between carriers and shippers. There are some barriers for the application of the information sharing system to city logistics. It is necessary to have lead-time of 2 or 3 days before dispatching of TL. This lead-time includes order processing and freight and fleet management for each truck carrier. And also the institutional elements such as the arrangement on the level of freight charge and the insurance on cargo and transport is very important pre-requites for information sharing. In near future, the possibility of application of information sharing system for city logistics will be increase when mobile Internet systems shorten the necessary lead -time and the institutional environment are improved.
422
Logistics systems for sustainable systems
INTRODUCTION Nation-wide transport information sharing systems using PC based online systems between trucking carriers have been in operation for the last decade in Japan. Two major networks exist, Network KIT and Local Net, with over one thousand carriers participating in each system. Since the announcement by the government in January 2000 regarding the development of a web-based marketplace between shippers and carriers as an information sharing system, various web-based transport exchange systems emerged. This paper aims to assess the present status and define the viability of such new information sharing systems and their impact on road traffic volumes.
STRUCTURE OF THE ROAD HAULAGE INDUSTRY
Market volume Japanese logistics costs including transport, warehousing and related administration costs are estimated at about US$ 374 billion in Fiscal Year 1999. This figure comprises about 8.8% of (he Gross Domestic Product (GDP) of US$4,256 billion, although the logistics cost to GDP ratio varies due to changes in the System of National Accounts (SNA). The share of transport cost is about 65.3% of the total logistics cost and is equivalent to about US$ 244 billion. The road haulage industry accounts for about US$ 92 billion (38%) and private trucks account for US$ 152 billion (62%) of the total transport cost (JILS, 2002). Recently, the government is aiming to increase international competitiveness of logistics activities. Transport is an important activity to decrease logistics costs.
Number of carriers and vehicles Freight transport is generally classified into two types: Truck Load (TL) and Less-than-Truck Load (LTL). There are 55,427 companies making up the road haulage industry in April 2002. Two hundred and seventy two (272) companies are classified as LTL type, 50,401 companies as TL type and the remainder classified as specialised type. The structure of trucking carriers is basically divided into two groups - one group consisting of large carriers, such as Nippon Jixpress which operate global freight transport services, and the other group which comprises 99% of trucking carriers, consists of Small and Medium Enterprises (SME). There are various trip patterns depending on the type of cargo and its delivery area. The number of freight vehicles is approximately 8.1 million^ and the number of vehicles belonging to the road haulage industry is about 1.1 million as of April 2002 (JTA, 2002).
Traffic volume and efficiency Freight vehicles comprise about 40% of the total road traffic when expressed in vehicle-km units. Trucks from the road haulage industry account for about half of the freight vehicle-kilometres. Private commercial trucks make the other half. Although private commercial trucks consist of the majority of freight vehicles-kilometres, they have relatively shorter trips and large number of vehicles. For trucks over 4 tons in 2000, the average distance for the road haulage truck is about 240 km per trip while the private truck averages about 93 km per trip (MLJT, 2000). Transport efficiency in terms of loading rate, which is the ratio of loaded ton-km to maximum loading
Web-based transport exchange systems in Japan
423
weight ton-km exclusive of empty vehicles, is 66.5% for road haulage trucks and 52.1% for private trucks. Vehicles from the road haulage industry are preferably used by non-specific shippers, such as common carriers (MLJT, 2000). Therefore, the loading rate of the road haulage industry is comparatively higher than private trucks. The loading rate of trucks over 4 tons including empty vehicles is 52.1 % for road haulage trucks and 27.8% for private tracks. For small tracks less than 4 tons, the loading rate is 29.7% for road haulage tracks and 11.6% for private trucks (MLTT, 2000). Therefore, private trucks can be considered as the main cause of traffic in the urban freight environment However, the detailed situation for private tracks is considerably more vague due to lack of data and its diverse operational characteristics, such as the variety of handling cargoes, diversity of trip patterns, multi-functional work of drivers, different receiving orders, and various allocation of goods. These activities normally result in long parking hours in front of retailer's shops.
METHOD OF TRANSPORT AND IMPACT ON TRAFFIC VOLUME Carriers generally have two ways to decrease road freight traffic volume. First, is to increase loading rates of LTL vehicles by joint delivery, and the second is to decrease trucks running empty through the use of a matching system that will allocate returning empty trucks to cargoes that need to be delivered along the way. Joint delivery is possible through improving consolidated services, collecting and delivery scheduling and routing methods. Matching systems are possible through the use of an information sharing system between carriers. Track load (TL) transport is usually one-directional from the consignor to the consignee. Thus, the returning vehicle generally has no cargo. For example, the transportation of fresh products such as vegetables and fruits from the rural area to the metropolitan area usually has high track loading rates. However, the returning transport routinely has empty cargo or contains few types of cargoes that basically depend on the demand of the population and industry in the rural area. Thus, the volume of cargo for the returning transport is relatively smaller than the outward transport However, it should be recognised that there are daily demands of goods from the metropolitan area to the rural area. This should be taken into account in the use of information-sharing and transport exchange systems.
424
Logistics systems for sustainable systems
Figure 1 Information sharing system and impact on traffic volume Figure 1 illustrates the basic concept of an information-sharing system and its effect on traffic. Without information sharing, Carriers A and B perform separate deliveries from/to Areas B and A. After delivery, both trucks do not have any return cargo and, thus, are forced to return back running empty. With an information sharing system in place between Carriers A and B, Carrier A can eliminate its empty trip if carrier B asks for its cargo to be transported by Carrier A. Carrier B will likewise be able to reduce its truck fleet This is one scheme that can lead to increased transport efficiency.
USE OF INFORMATION SYSTEMS BY THE ROAD HAULAGE INDUSTRY As mentioned earlier, most trucking carriers are SMEs. They do not have necessary resources to invest in and the ability to maintain information and communication systems (ICTs). Over the last 20 years, only large trucking carriers can afford to have on-line ordering systems with large shippers, inspection and tracking systems using bar code technology and sorting machines in truck terminals, and specialised mobile radio systems in delivery vehicles. Some carriers, which have truck fleets of about 100 to 300 vehicles, have utilised advanced in-vehicle units and fleet management systems. However, the infinite possibilities that the Internet and mobile phones provide have changed business operations.
Web-based transport exchange systems in Japan Mobile phone
Source: Japan Trucking Association, Report on ICT use in Trucking Carriers, March 2003 Figure 3 Percentage shares of mobile communications
425
426
Logistics systems for sustainable systems
Internet At the same time, the Intemet has become a popular means of communication for business activity (Figure 4). Thus, the difference between large and small companies has become insignificant
Source: Japanese Ministry of Public Management, Home Affairs, Posts and Telecommunications (MPHPT), Survey on the use of communication system in 2002
Figure 4 Percentage of Internet users by company size
Personal computer (PC) With the rapid diffusion of the Internet the use of Personal Computers (PCs) has also become widespread. In 1996, the percentage of PC users in the trucking industry was only 47.0%, of which half of the companies only had a single PC (Figure 5). Their use was limited to word processing and simple calculations using spreadsheet and other basic applications. However, in 2002, most of the trucking carriers have begun to use PC for Intemet and Web-based services. Source: Japan Trucking Association, Report on ICT use in Trucking Carriers, March 2003
Figure 5 Percentage of PC users in trucking companies
Web-based transport exchange systems in Japan
427
PRESENT STATUS OF THE INFORMATION SHARING SYSTEM A survey of the status of transportation information sharing systems in the Internet revealed that thirty-two (32) websites existed at the beginning of 2000. However, by May 2003, 8 websites have closed or have become non-active. At present, only 24 websites are active (Table 1). The survey also revealed that auction systems have been considered for inclusion in some of the websites. At present, there are no auction systems for transportation exchange in Japan. Although there were 3 websites that had bidding and automatic matching systems in 2000, at present, only one remains and has removed the auction system from its website. Table 1 Transportation Information Sharing Websites NAME URL Message board type between carriers http://www.jln.or.jp JL http://www.nikka-net.or.jp/kit/kit.html network KIT http://www.TraBox.com Tr@Box http://www.flamingo.gr.jp Flamingo e-carry http://www.n-o-k.com/e-carry/ Onimotu.com http://www.onimotsu.com/index.html Centre co-ordination type http://cweb.canon.jp/canoppvnet/ Canoppy Net http://www.jtplogi.co.jp jtp oneness http://www.oneness.ne.ip/on-frm.html QTIS http://www.krs.co.ip/tis/index.htm http://www.bsb-net.gr.jp/ Brigestone Loginet Ecologicom http://www.sti-corp.co.jp http://www.fujilogi.co.jp/unyu/index.html NEO ACTION Free Message board between shippers and carriers Planet Online Service http://www.plajion.co.jp Truck Gallery http://www.icho.com/truck/ jl09.com http://www.j 109.com/ http://www.f-logi.com/ f-logi http://www.yuix.co.jp/ YUIX Message board between shippers and carriers truck-net http://truck-net.co.jp/indexx.html http://www.i-tone.ne.jp/ i-tone Kaeribin Truck Kuusha Jyoho http://www.udl.co.ip/ The physical distribution 21 http://www.moritakonpou.co.jp/webnet NQQ net http://www.nkk-butsuryu.co.jp/product/nqq.html Bidding and auto matching | logilink http://www.j-logilink.com/
YEAR May,2000 Oct.,1991 Nov.,1999 June, 1993
July,2000 June,2000 June,2000 March, 1998 July, 2000 Jan.,2000 May,2000 Jan.,2000
June, 2000 Jan.,2001 April,2001
The survey of websites also revealed the following: (a) Out of the total 24 websites, 5 are of free type and 19 are of agent type when deciding transport rates (b) Out of the total 24 websites, 16 use message boards, 7 use agents, and only one uses computer aided matching for the matching process (c) Successful websites are those that use closed agents, including financial transaction based on insurance
428
Logistics systems for sustainable systems
systems or free message board type especially between carriers.
BUSINESS FLOW AND INFORMATION SHARING SYSTEMS It is important for information sharing systems to maintain a certain lead-time from the request of cargo or a vehicle between a specific origin and destination in order to respond to the demand. The lead-time depends on the business process. The standard flow of business for a TL carrier is as follows: 1) Order entry The shipper requests the carrier to transport its cargo under a specified set of conditions, and supplies the details of the consignor, consignee, and type of cargo and vehicle. 2) Scheduling and Planning The carrier makes scheduling and fleet management plans, such as the allocation of specific cargo to specific vehicles, selection of routes, vehicle type, driver, and departure/arrival time, and the determination of handling conditions for loading and unloading. 3) Loading This is the preparatory activity before actual transport, in which the cargo is loaded onto the vehicle in the carrier's terminal or at the consignor's depot. 4) Delivery This is the actual execution of transport After delivery, drivers need to report their driving records. In urban delivery, order entry occurs in the evening, cargo loading happens at midnight and cargo delivery takes place the following morning. Thus, there is no sufficient lead time to arrange and co-ordinate the cargo and vehicle with other carriers. In contrast, long haul transport usually needs overnight transport and rest time, and that it normally takes around 2 or 3 days to travel a distance of 600 to 1,500 kilometres. One round trip thus takes one week. This means that there is enough lead-time, but on the other hand, there may be difficulty in finding a fixed return cargo in a specific direction due to unstable production and demand conditions. Therefore, carriers need to find regular shippers who have need for a transport service for their return trips. Most information sharing systems have additional function that introduces trusted carriers in particular areas.
ROLE OF INFORMATION SHARING SYSTEMS The most important role of a information sharing system is to find a suitable partner for the shipper or carrier requesting or offering transportation services. Thus, one of its primary objectives is to introduce to the shipper a trusted carrier who is able to perform quality and dependable long distance transport The system must perform this function at all times. After an appropriate partner or carrier is selected, the carriers usually contact the shipper directly by telephone. Web-based information sharing systems have a wide variety of functions ranging from message boards to financial transactions and vehicle or cargo location monitoring (Figure 6). However, the success of
Web-based transport exchange systems in Japan
429
information sharing systems depends on the willingness of carriers and shippers to utilise the system. The following paragraph, which describes the apprehension towards the use of information sharing systems, also applies in Japan: "For a shipper, price is not often a defining factor, quality of service and dependability is! Auction models were of value to shippers only in cost savings, and this was often associated with extra risk which comes about from using untested/unknown transportation providers, the cost savings did not outweigh the risks involved...plus the fact that contract transportation makes up 90% of the market resulted in a shipper receiving better rates only on a small proportion of its logistics requirements for an increased risk. The other great barrier to the auction model was from the carriers...a lot of them simply did not want to play" (http://www.eyefortransport.com/archive/ newslettered29.shtml). The other significant issue concerning the use of information sharing systems is whether there is a legal structure or institutional framework in which the system can operate. The future viability of these systems will depend on the institutional environment and the development of ICT applications. With these systems in place, increased information sharing which will have significant results on the improvement of city logistics is possible.
Figure 6 Sample display of a matching system
CONCLUSION Web-based business transactions at present should only be designated for specific players and should not be available to the entire marketplace in order to be successful. Thus, its impact in decreasing traffic volume due to reduced empty vehicles is limited at this point in time. Although its main application is in long distance freight transport, it is expected in the near future that information exchange for short and medium distance transport using Internet accessible mobile phones with Global Positioning System (GPS) or
430
Logistics systems for sustainable systems
wireless data communication device will increase and function as a new tool for increasing transport efficiency.
REFERENCES Japan Institute for Logistics System (JILS), Survey on Logistics Cost, (2002). JILS, Tokyo, http7/www.logistics.or.jp/jils/english/english.htm Japan Trucking Association (JTA), White Paper on Trucking Industry (2002) JTA, Tokyo http://www.jta.or.jp (Japanese only) Japan Trucking Association (JTA), Report on ICT use in Trucking Carriers, (2003). JTA, Tokyo, http://www.jta.or.jp(Japanese only) Japanese Ministry of Land, Infrastructure and Transport (MLJT), Census on Road Traffic, (2000). MLTT, Tokyo, http://www.rnlitgo.jp/englishAidexJitml Japanese Ministry of Public Management, Home Affairs, Posts and Telecommunications (MPHPT), White Paper, (2002). MPHPT, Tokyojittp://www.soumu.go.jp/joho_tsusin/eng/index.htm Junichi Fujimaki, Seicho Tojo no Kyuka Kyusha System, Keiei Joho Research, Daiwa Research Insu'tute(DRr), 1 June 2001, DRL TokyoJ(ittp://www.dir.co.jp/kj/Search/Search_013202.html (in Japanese). Masahiro Oya, Kyuusha Kyuuka System, Ryutsuu Sekkei, Aug.2000, Yusou Keizai Shinbunsha, Tokyo, http://www.yuso.co.jp/ (in Japanese) E-commerce for Freight and Transport Newsletter, Edition 29,13 Dec 2000, http://www.eyefortransportcom/arcMve/newsletteiied29.shlrrd
31 SUMMARY OF THE OECD REPORT DELIVERING THE GOODS - 21 ST CENTURY CHALLENGES TO URBAN GOODS TRANSPORT'
OECD Programme of Research on Road Transport and Intermodal Linkages Working Group on Urban Freight Logistics Chairman: Fred J.P.Heuer, drs. the Netherlands
ABSTRACT Although delivery of goods is vitally important for residents and industries in urban areas, the presence and operations of goods transport vehicles in urban areas are often regarded more as a nuisance than an essential service. Relatively little has been done by many governments to facilitate the essential flows of goods in urban areas and to reduce the adverse impacts of urban goods transport on the communities being served. This has resulted in increasing problems associated with goods delivery including competition with passenger transport for access to road infrastructure and space for parking/delivery facilities. How should OECD countries deal with the difficult challenges they face in this area? This report analyses measures taken in many cities in the OECD area and provides recommendations for dealing with these challenges.
INTRODUCTION Goods transport in urban areas has a major impact on the economic power, quality of life, accessibility and attractiveness of the local community, but it receives little attention in comparison to passenger movement. With the ongoing increase in urban goods transport, there is increased concern about goods movement and its consequences. There are many solutions that have been proposed and implemented in OECD Member countries with both successes and failures.
432
Logistics systems for sustainable cities
The Working Group on Urban Freight Logistics was set up to learn from such international experiences with the aim of identifying what could improve the efficiency of urban goods transport systems, while ensuring the environmental sustainability and liveability of urban areas. The members of the Working Group gathered information on urban freight policies from different OECD countries. The summary of the report is presented below. This is the first OECD report that is fully devoted to the topic of the delivery of goods in urban areas. The limited and fragmented information available in this area meant that some aspects could not be addressed as comprehensively as others.
OVERVIEW Definition of urban freight transport For the purposes of this report, the Working Group focused on the delivery of consumer goods and defined urban goods transport as: "The delivery of consumer goods (not only by retail, but also by other sectors such as manufacturing) in city and suburban areas, including the reverse flow of used goods in terms of clean waste ". The Working Group recognised that delivery of consumer goods is only part of the whole logistics chain and should therefore be considered from a broader systems perspective. Consequently, the report provides a more encompassing view of urban freight logistics and its problems. Developments in society and policy making Urban goods transport issues are a result of a wide pattern of developments in our society, which include moving towards a post-industrial society, ageing and individualisation of society, urbanisation, as well as sustainable development which is becoming the guiding vision for many OECD countries. Policy making in such a society requires well-designed consultation and participation processes, due to the complexity of issues involved and diverse interests among various stakeholders. This is particularly the case for policy making for urban goods transport, since it involves many different parties with separate and often conflicting interests, who have to share limited urban space. The complex operations of urban goods transport and the variety of problems caused thereby further complicate policy making in this area.
Developments in freight transport Freight transport is a fundamental component of urban life. Globalisation of economic activities, changes in consumer behaviour and developments in advanced technologies have led to many developments in freight transport. Businesses have expanded the area of their sourcing and distribution operations, developing world-wide supply chains that link customers, suppliers and manufacturers. Urban goods transport has therefore become integrated with long-haul transport. Businesses seek to improve the flow of their supply chains by utilising Information and
Delivering the goods-21st century challenges to urban goods transport
433
Communication Technologies (ICT) and optimise such supply chains by reducing the number of warehouses, centralising inventory and consolidating their deliveries. The retail sector seeks to minimise costs by saving storage space and reducing stock, resulting in strict demands being placed on the supply chain which include reduced delivery lead times and just-in-time deliveries. As customers become increasingly integrated in the supply chain, the need to respond more rapidly to varied and often-changing customer demand requires the flow of the supply chain to be increasingly time sensitive. The rapid development of E-commerce also requires fast and reliable delivery. These developments have led to increases in freight transport and further increases are unavoidable if no additional measures are taken. However, the various negative impacts show that the impact of continued growth in freight transport is not sustainable in the long term. Therefore, efficient organisation of urban goods transport has become crucial not only for successful supply chain management and the development of E-commerce, but also for sustainable development. The demand for just-in-time, tailor-made urban goods deliveries, which is difficult for nonroad modes to meet, poses a challenge to the development of intermodal transport, although considerable efforts are being made in some countries to find intermodal solutions. Urban goods delivery by road transport raises another issue: the type of vehicles to be used. Smaller vehicles are often used for deliveries in urban areas, although they tend to generate more traffic and energy inefficiencies than heavier trucks. Increases in the number of passenger vehicles have led to passenger and freight transport competing for limited urban space, with the former often receiving priority in policy-making. Passenger vehicles can be a final link in the logistic chain, since people make use of cars to bring goods to their homes. Urban goods transport policies need to take into consideration the interactions between passenger and goods traffic.
Problems of urban goods transport Since urban goods transport takes place in areas with a high density of population and mixed use of public space, various problems have been encountered in many cities. Accessibility problems are both encountered and caused by urban goods transport. Problems encountered by freight vehicles are mainly due to insufficient infrastructure, access restrictions or congestion. This results in freight vehicles causing disruption to traffic and further congestion. Freight transport contributes considerably to environmental problems such as emissions, noise, vibration and physical hindrance. It also causes safety problems since freight vehicles, due to their size, manoeuvrability and on-road loading/unloading operations, are a significant cause of accidents. Urban goods transport is a major and rapidly growing sector of oil consumption, which gives rise to problems of energy consumption and related emissions concerns.
434
Logistics systems for sustainable cities
These problems have led to some increased concerns about the consequences of urban goods transport. Although it is clear that urban goods transport is crucial for maintaining the economic and social functioning of cities, there seems to be a serious lack of awareness of its benefits. Awareness of urban goods transport seems to be rather one-sided, focusing more on its problems than on its importance.
LESSONS LEARNED FROM APPROACHES IN MEMBER COUNTRIES Countries are in different phases concerning policy development on urban goods transport. However, from the experiences in member countries, the following lessons can be learned. Different situations, common challenges While being increasingly concerned about negative impacts of urban goods transport, cities are aware that delivering goods to the city is essential for maintaining their economic and social functions. Therefore, cities are confronted with common and difficult challenges of maintaining their sustainability and liveability while ensuring a goods transport system that sufficiently serves their needs. The extent of national government involvement in urban goods transport varies In many countries, problems of urban goods transport are dealt with at a local or regional level, resulting in lack of consistency among local or regional measures. Only a few countries have developed an explicit encompassing national policy focused on urban goods transport. Lack of awareness and knowledge is a serious obstacle There is a lack of awareness and knowledge of urban goods transport not only among the general public but also among governments and city planners. This has often led to transport related policies and facilities being planned merely from the passenger transport perspective, without adequate consideration of the needs of freight transport. Lack ofbefore-and-qfter evaluations and data Few countries have analytical tools and data for evaluating the effectiveness of their policy measures concerning urban goods transport, resulting in their measures causing unexpected side-effects. Policy measures tend to lack long-term and supply-chain perspectives Policies currently in place tend to focus strongly on short-term problems and solutions. Few attempts seem to have been made to provide forecasts for future developments or to develop long-term policy options. Also, in spite of the fact that urban goods transport is integrated with long distance transport, current measures on urban goods transport often only take account of the urban area and pay little attention to the supply chain as a whole. Regulations tend to be not harmonised, unstable and often lack enforcement Local regulations tend to differ among different municipalities and be changed as circumstances change. This can cause difficulty in enforcing such regulations on drivers who are often not aware of the different and changing restrictions. Such a lack of harmonisation and stability also causes problems for the vehicle manufacturing industry in developing vehicles that comply with such regulations.
Delivering the goods-21st century challenges to urban goods transport
435
Public-private platforms seem to be helpful Since urban goods transport issues are complex and involve many stakeholders, consultation platforms have proved to work well in some countries in bringing such stakeholders together for discussing issues and planning measures. Non-market based urban distribution centres tend to fail Publicly owned or publicly-driven distribution centres often do not receive support from the private sector and become commercially unsuccessful. Consolidation seems to be an emerging trend Consolidation of deliveries is emerging as an important tool for solving problems, but little attention is paid to accommodating or facilitating this through policy measures. Innovative policies are being attempted Some countries are attempting to implement innovative policy measures, e.g., selective timesharing and multiple use of infrastructure, introducing environmental zones and using pricing for diverting freight traffic from residential areas, with some promising results.
POLICY RECOMMENDATIONS Urban goods transport is now facing many difficult challenges. However, the opportunities for dealing with such challenges have increased in recent years, since the civil society has become aware of the need for sustainable development and is realising that it is a common responsibility of both public and private actors. Experiences show that single-shot measures, planned and implemented by local governments alone, are generally not sufficient in developing a sustainable urban goods transport system. Therefore, consideration has been given to the policy framework necessary for developing such a system as well as recommendations on actual measures. Policy framework National/State government initiatives are crucial In order to apply consistent, stable and effective measures throughout the supply chain, National/State governments need to take the initiative and provide clear policy objectives and frameworks under which tailor-made local measures can be planned and implemented. The main policy objective should be "sustainable urban goods transport" Continuing economic growth while protecting the environment and ensuring a better quality of life for future generations are foremost objectives in OECD Member countries. Therefore, the main national objective should be sustainable urban goods transport, which requires the development of an urban goods transport system on a socially, economically and environmentally sound basis. Both short and long-term policies should be developed under this objective. Urban goods transport policy needs consultative planning - importance ofPublic-Private Partnerships Urban goods transport involves a wide range of public and private actors, with different and often conflicting interests, who interact interdependently. Agreement among all stakeholders,
436
Logistics systems for sustainable cities
especially support from the private sector, is necessary in developing a feasible and practical policy vision. Therefore, consultation can be considered to be a major part of establishing a sound policy framework for urban goods transport. Since urban goods transport has become a final leg in global supply chains, the actual consultation process needs a supply chain management perspective, with the involvement of stakeholders responsible for national or international supply chains. Public-Private Partnerships - where various government layers, shippers, transport operators, vehicle manufacturers, retail and wholesale organisations, real estate developers, research bodies and inhabitants all co-operate closely in developing common objectives and solutions - are necessary for effective action. Integration of policies and measures across sectors is important Since transport and logistics are interrelated with international trade and regional and local concerns, the policy framework should be seen in a much broader context. Integrating policies not only with passenger transport but also among different policy areas and different levels of government is necessary for establishing a more effective urban goods transport policy. Policies should be formulated so as to enhance developments in the private sector The private sector has become increasingly aware of its roles and responsibilities and is active in developing sustainable urban goods transport systems. Many developments in increasing efficiency and reducing negative impacts of urban goods transport systems are initiated by the private sector. Policy measures should be formulated so as to enhance and facilitate such developments. Regulations need to be sufficiently harmonised and stable so as to provide a clear framework to encourage the private sector to assess the effectiveness and viability of potential investments. Planning through a Public-Private Partnership process can guarantee that the measures are practical and that the private sector is committed to such measures. Active and continuous campaigning, including promulgating best practice, is also important in order to stimulate and foster the awareness of the private sector.
Recommendations on measures - dealing with new challenges Drawing on experiences in member countries, the following non-prioritised recommendations are proposed in implementing such measures within the proposed framework. 1. Active measures are needed to increase awareness of the importance of urban goods transport and to diffuse knowledge. Increasing awareness and knowledge in urban goods transport is a starting point for developing an efficient goods transport system. Governments should encourage public awareness on the importance of urban goods transport in their daily lives, the progress made so far, and the future challenges concerning urban goods transport which require participation of all stakeholders in order to be resolved. Communication and consultation processes using PublicPrivate Partnerships can be useful to increase such awareness and diffuse knowledge among all stakeholders. In order to diffuse and increase awareness and knowledge in local governments, one useful procedure may be for the National or State governments to require local governments to
Delivering the goods-21st century challenges to urban goods transport 437 formulate local transport plans that include urban freight transport and have local government consult National/State governments on the plans. Local governments will be compelled to increase awareness of urban goods transport issues and their knowledge will improve accordingly. This will also contribute to achieving consistency among local measures. In the initial phase, it will be useful for the National/State governments to provide guidance and consultation to local governments. 2. Evaluation methods and data are prerequisites for effective policy measures In order to plan and implement effective urban goods transport policies, both before and after (ex-ante and ex-post) evaluation methods need to be used from the planning phase through to their implementation. All stakeholders need to reach a consensus on clear policy objectives, indicators to measure the achievement of such objectives, and a standardised evaluation method for planning and monitoring the effectiveness of measures actually taken, using the agreed indicators. National/State governments should encourage local governments to implement both ex-ante and ex-post evaluations. In relation to planning vehicle access and freight traffic restrictions, it would be desirable for possible regulations to be evaluated, including for cost effectiveness, prior to their adoption and implementation. Ex-post evaluation is also necessary, both for monitoring and benchmarking measures, and for comparing the results with the ex-ante estimates, thereby improving the evaluation method. Data necessary for evaluation methods should be collected in a consistent manner with sufficient standardisation so as to allow long-term monitoring and benchmarking. For this, agreement on the definition and collecting methods for all data needs to be reached, preferably on an international basis. 3. Consolidation is a key to achieving sustainable urban goods transport With increasing demands for frequent and just-in-time delivery on one hand and restrictions of limited spatial infrastructure and environmental demands on the other, future solutions for achieving sustainable urban goods transport should be sought through the consolidation of goods delivery. The purpose of consolidation is to improve the utilisation of the transport system to generate economies of scale, thereby reducing vehicle trips, increasing efficiency and decreasing financial and environmental costs of transport. A useful measure for improving consolidation is the implementation of a commercial urban transhipment centre, where freight destined for the urban area would be sorted into consolidated loads for final delivery. Community collection and delivery points could also be used to improve goods consolidation. As consolidated loads generally would be delivered by small vehicles, the highest possible vehicle utilisation is necessary in order to compensate for the additional transhipment cost and to ensure reduction of vehicle kilometres. Using ICT for managing available capacity, making optimal vehicle utilisation and route planning could achieve this. Although consolidation has been mainly driven by the private sector in the form of voluntary co-operation, governments are able to promote such consolidation by way of encouraging and assisting pilot projects and by favourable regulations. 4. Regulations need to be harmonised, standardised, stable, easy to enforce and costeffective Various regulations have been implemented that aim to maintain the living environment in certain urban areas and to facilitate smooth and safe traffic flows. Of these regulations, access
438
Logistics systems for sustainable cities
restrictions based on time and/or vehicle size or weight have been widely implemented, especially in Europe. Such restrictions differ among municipalities and are often not sufficiently explained to drivers, causing serious difficulties for operators who organise worldwide supply chains. In order to achieve transparency as well as stability in long term policies, it is important that better harmonisation be achieved on truck size and weight definitions. Existing regulations on truck size and weight should be reviewed for consistency, if possible, making them simpler and closer to the professional needs of carriers, shippers and retailers. Such reviews need to be promoted by both national government initiatives and by international co-operation, while ensuring private participation in the decision making process. Regulations related to transport vehicles are crucial for vehicle manufacturing industry and fleet owners. As wide as possible standardisation of clear regulations applied for a sufficiently long period is necessary to encourage vehicle manufacturers to develop low-noise and lowemission delivery vehicles. Ideally, a limited number of recommended "ideal size" of truck dimension limits for city access should be determined internationally. Harmonisation and standardisation of regulations related to vehicles can also facilitate the consolidation of goods between shippers and transporters. Enforcement is always an important issue. Regulations should be designed in such a way (clear, simple, easy to understand, cost-effective, and where relevant, preferably performance based) that they are easy to enforce. Lack of control and enforcement has made policies less effective, resulting in regulations often being ignored, especially by passenger vehicles using infrastructure provided for freight vehicles. Strong control and enforcement is necessary and has been made possible, due to the development of new monitoring techniques and tools. 5. Infrastructure capacity should be used more imaginatively on a 24-hour basis In order to make optimal use of the limited urban infrastructure while maintaining accessibility and liveability in cities, selective allocation of infrastructure capacity on a 24- hour basis needs to be considered. Such allocation schemes serve to separate infrastructure use in terms of time and space per type of vehicle based on their characteristics. Experiments in mixed use of streets have proved satisfactory and have shown that acceptance by all stakeholders and effective enforcement are crucial for their success. An important measure under discussion is the introduction of night deliveries. Although freight vehicles have been banned from many urban areas during night time due to their noise problems, studies show that night deliveries could reduce the concentration of activities and road congestion during the day, resulting in removing traffic from peak hours and improving efficiency of deliveries, which in turn produce cost and environmental benefits. In order to become acceptable, night deliveries need to be considered in conjunction with the development of quieter delivery operations - including quieter vehicles and loading/unloading facilities. Innovative vehicles and equipment need to be developed and experimented. Governments can promote such developments by favourable regulations or financial incentives. Consultation is necessary to achieve acceptance by the local community. Experiments and pilot projects are useful during the consultation process, since these enable residents to experience actual low-level noise operations before agreeing to a formal change of regulations being introduced.
Delivering the goods-21st century challenges to urban goods transport
439
6 Cleaner, low noise and more energy-efficient vehicles need to be promoted Innovation in vehicles including environmentally friendly and energy efficient engines, onboard routing systems, delivery-suited vehicle designs, and delivery handling equipment should be promoted by providing incentives, providing information and by establishing clear and stable international standards. 7. Adequate logistic facilities need to be provided In order to increase the efficiency of urban goods transport and at the same time reduce negative impacts concerning road use, loading/unloading zones need to be provided. Location and time-periods of such zones should be carefully planned to accommodate freight vehicle operations in a most efficient manner, clearly signed and their use strictly enforced. Off-road loading/unloading facilities for new buildings should be included in zoning codes and building permit requirements. Co-ordinated actions should be pursued with the private sector to develop transhipment facilities and facilities for home delivery which will contribute to the consolidation of goods, utilisation of intermodal transport, and efficient home deliveries. 8. Efforts need to be made to reduce safety risks of urban goods transport The often severe consequences of accidents involving goods deliveries have greatly contributed to the negative image of freight vehicles. Governments should provide the necessary infrastructure, with private participation where appropriate, to reduce risks of accidents involving freight vehicles. Governments also need to strengthen their control regarding freight transport operations and promote efforts by the private sector to reduce safety risks of their operations. 9. Reverse logistics needs to be developed The imminent need in many countries to reduce, reuse and recycle waste will only become feasible with a transport system which carries used and returned goods for reuse and recycling (reversed logistics) in a cost effective manner. Governments can facilitate the development of efficient reverse logistic systems by providing necessary infrastructure and by diffusing and encouraging best practice.
10. Technological and conceptual innovation can support sustainable urban goods transport Various measures for developing sustainable urban goods transport systems will be made possible due to innovative technology development. For governments, with the use of ICT, flexible implementation of access restrictions, loading/unloading zones, and transport demand management schemes will become feasible and easily enforceable. The evolution of City Logistics also offers opportunities for this concept. Development of underground distribution systems offers possibilities for more sustainable urban goods transport systems, but would require an active government role. Technology developments in the private sector also contribute to increasing efficiency and reducing cost and environmental impacts, and therefore should be promoted by facilitating experiments and diffusing best practice.
440
Logistics systems for sustainable cities
11. Next steps - Needfurther study and international co-operation It became clear during the studies by the Working Group that the development of sustainable urban goods transport is only in its initial phase. Further studies and collection of data are clearly necessary. However, it was encouraging to find that countries have begun to be aware of the importance and problems of urban goods transport, and are trying out various measures to meet the challenges. Since such challenges are common among most countries, it proved to be extremely useful to exchange experiences amongst the countries. Further international cooperation is necessary not only in sharing best practices, but also in harmonising regulations, standards and data collection.
32
SUSTAINABLE CITY LOGISTIC SOLUTIONS
Soeren Kjaersgaard, City of Copenhagen, Denmark Henrik Enslev Jensen, City of Copenhagen, Denmark
ABSTRACT Sustainable city logistics solutions must be considered in accordance to develop lively and accessible city centres where we can all move around safely, and where trade and culture are flourishing. This requires delivery of goods on a daily basis considering these aspects: • • •
The transport must be geographically concentrated, Large amounts of volume of parcels or goods High exploitation of capacity
Copenhagen - certification scheme From 2002 to 2003 all vans and lorries over 2,5 tons are required to have a certificate in order to stop in the medieval heart of Copenhagen. The compulsory certification scheme includes requirements of capacity utilization and engine technology, and should reduce the number and/or size of vehicles and lower the visual intrusion from street traffic. This scheme is the first project acknowledged by the Danish Ministry of Transport under the environmental zone scheme. This compulsory trial is designed from a l'/i years voluntary trial with 80 transport companies.
Results It is difficult to register any significant effect of the ordinance in decrease of traffic congestion or less exhaust. The main part of cars between 2,5 and 18 tons stopping in the Medieval centre has a certificate and follows as such the restriction. Many of those cars though have a red certificate without demands to the car or capacity utilization. Furthermore a big part of the cars
442
Logistics systems for sustainable cities
are drive through. Traffic censuses shows that the total amount of trips by vans and lorries in 2003 still are at the level of 1999. In the same period the number of small vans between 2 and 2,5 tons has increased.
INTRODUCTION The City Goods Ordinance is an obligatory Pilot Scheme lasting from 1. February 2002 to 31. October 2003. The name City Goods means city cargo, aiming to the key elements of the ordinance. This means, that vans and lorries between 2.500 kg and 18.000 kg total weight must be 60% full/loaded and have an engine less than 8 years old, if they want to enter the innermost part of Copenhagen. This part of the city is called he Medieval City, because of its narrow streets dating back from the medieval times, and it's l x l km in area. Today, based on more than one year with the City Goods ordinance, it has been clear, that the traffic demand related to craftsmen, couriers and other services are much bigger than first presumed. Therefore new rules for this traffic have been implemented to insure, that these cars also participate in reducing the environmental pressure on the medieval city. Trucks heavier than 18 tons must have special permission to enter the area. The main objective of the Copenhagen City Goods ordinance is to increase the use of capacity of the vans and lorries driving into the Medieval Copenhagen. The philosophy behind the ordinance is, that a reduction in the number and/or size of lorries and vans are mainly to be met by a demand of maximizing the use of the capacity rather than - or as a supplement to demanding cleaner engines or setting up toll zones.
BACKGROUND A report from 1996 says, that about 3,500 vans and lorries made approximately 6.000 trips in and out of the Medieval city of Copenhagen every day, making it difficult to navigate around, creating pollution and making people in the area feel insecure. Furthermore there is an increase in the damage of street inventory (signs and so on). The report also showed a very poor utilization of capacity among trucks and lorries. In 1996, a working group, consisting 5 of the biggest transporters (Carlsberg, the Danish mail, etc), some hauliers organizations and the municipal of Copenhagen, was established, to come up with solutions dealing with the problem. The work resulted, after more than 20 meetings, in the City Goods ordinance, which was introduced in 1998 as a voluntary test scheme that ran for 1 1/2 years until January 2000. 80 companies participated in the scheme with 350 vans and lorries. In 2000 a small adjustment of the Danish traffic-law made it possible to perform obligatory test schemes due to environmental reasons. Based on the new law, the full-scale obligatory test Scheme started in Copenhagen in February 2002 using many of the experiences from the voluntary test scheme.
Sustainable city logistic solutions
443
THE CITY GOODS ORDINANCE To reach the goal, vans and lorries over 2,500 kg total weight must have a City Goods certificate to stop legally in the medieval city centre. The municipal traffic warden controls the rules, and violators can be fined 510 DKr (app. 68 EURO). Types of certificates Green The green certificate is the main certificate for the actual cargo transporters, who on average utilize the capacity of every single vehicle with 60% over a 3-month period, and uses a vehicle with an engine younger than 8 years. Once every 3 months the company must send a report of the capacity use to the municipality. Each transporter reports the utilization due to their own application form, thus they can calculate the freight in kilograms, liters, pallets or other units. The green certificate gives exclusive rights to use the 26 special loading zones (Monday-Friday, 8-12) that are established around the area in connection with the City Goods ordinance. To obtain a certificate the owner must apply for the certificate on a special application form. The certificate costs 260 DKr (app. 37 EURO) and is valid for the entire trial period. Yellow
During the first year within the trial scheme the yellow certificate was an option for those who could not meet the Green Certificate's restrictions, thus it functioned as an amnesty. After adjusting the rules due to the experiences reached in the first year, it nowadays functions as a certificate for craftsmen, couriers and other services driving in vehicles not longer than 6 meters long (external measurement) or the vehicle has a maximum permissible total weight of 3.5 tons and with an engine not older than from the Is' of January 1997. Also the yellow certificate must be applied for on a special application form, and costs 260 DKr (approximately 37 euros) and is valid for 1/2 a year. Vehicles that belong to a company on a special branch code list approved by the City Goods secretary can have an engine not older than from 1st of January 1995. Red The red certificate is an option for those who rarely come to the Copenhagen inner city, and for cars that cannot meet the demands for either a green or yellow certificate. The certificate can be bought without application form around the clock at app. 25 petrol stations on streets leading to the designated area. The certificate costs 40 DKr (approximately 6 euros) and is valid for only one day.
INFORMATION TO THE TRANSPORTERS Prior to the introduction of the City Goods Ordinance, an information campaign was held in newspapers and trade journals with articles and advertisements. The approximately 40 roads leading into the medieval area regulated by the City Goods ordinance are marked with traffic signs which also mark the 26 loading zones. Due to the fact that the certificate only allows stopping, not parking, there are also signs on the 110 ticket machines in the area, so that people buying parking time are informed about the ordinance.
444
Logistics systems for sustainable cities
CONTROL OF THE ORDINANCE The control is performed in two ways, on the streets and administrative. The Copenhagen Municipality traffic warden gives fines to vans and lorries that have stopped in the streets of the Medieval city without having a City Goods certificate. Likewise they fine all cars, parking in one of the 26 loading zones, between 8-12, if they do not hold a green certificate. Those holding a green certificate, must every 3 months send in a report on the capacity utilization, to the City Goods administration. The report must have information on the amount of cargo on every trip into the designated area (transit goods must be excluded). Among those having a capacity utilization exceeding 60%, the City Goods administration will perform spot-checks, asking the transporters to verify the information given.
RESULTS AND EFFECTS As a status report on the City Goods Ordinance has just been made, it's possible to give new data on the results of the scheme by June 2003. The report is based on hard and soft data collected from traffic censuses, data collected from the applications in the City Goods database and from questionnaires amongst inhabitants and businesses in the medieval city and amongst the transporters and car owners.
Traffic censuses From the traffic censuses we can conclude, that the overall traffic by cargo cars in the medieval city from 1999 to April 2003 are close to status quo (Table 1). It also showed that about half of the cars registered in the censuses have a City Goods Certificate. The only significant change is the growth of cars under 2.5 t and the average tonnage falling from 4.5 to 4.2 t, thus the result shows a movement towards usage of smaller cars. The census held by 5. August 2003 shows a decrease in the total registrations, probably due to a general seasonal decrease in traffic at about 10 %. Table 1 Results from four traffic censuses 1999
2002
April 2003
August 2003
Total registrations
6.032
6.236
6.209
5.458
Total number of vehicles
3.543
3.503
3.979
3.501
Total number of vehicles above 2000 kg
3.117
3.155
3.526
3.144
Total number of vehicles above 2500 kg
2.456
2.153
2.341
2.185
10.770.100
9.776.200
11.161.325
9.979.375
4.385
4.540
4.767
4.567
Total Tonnage, kg. Vehicles above 2500 kg Average, kg
Certificates Yellow certificates are issued for periods of 3, 6 and 9 months, while green certificates were issued for the whole trial period under the condition that the reports on capacity utilization are
Sustainable city logistic solutions
445
given every third month. All together 3,600 certificates have been issued to 2,400 companies (Table 2). During the first 15 months of the trial period there has been a certain development between the yellow and the green certified cars, mainly because the rules have been changed. In the first year of certification it was possible to obtain an amnesty, meaning that it was possible to get yellow certificate to all cars between 2.5 and 18 tons. After 1st of February 2003 only vans fulfilling the specifications could get a certificate. Table 2 Number of issued certificates and companies with certificates in the three periods Number of:
Certificates Companies Green Yellow Companies
1. Period (1st half 2002) 2. Period (2nd half2002) 3. Period (2003)
853 942 909
2983 2637 2016
1967 2234 2381
Approximately 900 cars have had the green certificate. There have been some changes among the certificate holders, as some had lost their certificate due to violations of the conditions i.e. failing to give in the capacity reports. At the same time some trucks have moved to the green certificate, as they could no longer obtain the yellow certificate. The weight division for the lorries with green certificate in Table 3. Table 3 Number of issued certificates in the three periods due to weight classes for the cars Certificates 1. certificate period (1st half 2002) G re en Total w e i g h t 2501 - 2800 2801 - 3000 3001 - 3200 3201 - 3500 3501 - 6000 6001 - 12000 12001 - 18000
Total
46 14 175 309 58 76 175 853
2. certificate period (2nd half 2002) G re en
3. certificate pe riod (2003) Green
43 15 176 342 67 104 195
34 17 138 309 78 119 214
942
909
All the cars driving with the green certificate have made more than 16,000 trips to the area between 1st November 2002 and 31st January 2003, or divided into 5 working days a week about 250 trips a day (or 2 -3 trips per car a week). The approximate utilization among vans and lorries are reported to be close to 70% (Table 4). Many transport companies have due to the City Goods ordinance, maximized the capacity utilization by analyzing their logistics, by dialogs with their customers and by using the most suitable car for their task.
446
Logistics systems for sustainable cities Table 4 Reports on utilization for cars with green certificate for one three month period Number of reports Total weight 2501 - 2800 2801 - 3000 3001 - 3200 3201 - 3500 3501 - 6000 6001 - 12000 12001 - 18000
28 15 120 257 70 83 165
Total
738
Number of cars with trips to the zone 17 8 74 182 43 54 109 487
Average Number number of trips of trips 1085 328 2108 6169 1957 1485 3154
63.82 41.00 28.49 33.90 45.51 27.50 28.94
16286
38.45
Average utilisation % 65.53 84.75 66.86 70.47 65.37 73.11 64.81 70.13
Confiscated due to lack of reporting 4 0 21 35 3 8 8
79
On average 100 red certificates were sold per working day through the period, a total of approximately 35.000 certificates. It is not possible to tell anything about who are using the certificates, as they are sold without demands from 25 petrol stations. As mentioned, the traffic wardens have been issuing fines to those violating the rules, and they have issued approximately 13.500 fines by the end of the scheme. Issued parking fines due to City Goods
Figure 1 Number of Issued parking fines due to City Goods Questionnaires The reactions from the citizens and business in the affected area shows, that the main reactions are positive, and that inhabitants in the area feel an effect from the ordinance. The handicrafts in general have the biggest resistance against any regulation, whereas the transport workers are satisfied especially with the loading zone that has been made as a part of this trial. Another general issue is, that traffic regulations in general are welcome, but that specific groups can not see, that they should be affected by restrictions.
Sustainable city logistic solutions
447
Amongst the transporters covered by the ordinance, there is in general a more positive attitude towards regulations relating to the motor technologies than to the capacity requirements. Half of the respondents replied, that they had not noticed any relief of their daily work. About 10% of the transporters said, that it had made their work easier. Many holders of green certificates found that the reporting was difficult. Many found that the ordinance was only an extra tax and that companies themselves try to optimize the transport as it would also be the most profitable.
Other facts The City Goods Ordinance is a joint venture between Road & Park (planning) and Parking Copenhagen (operating). A secretary has been established in order to administer the ordinance. 3-4 people have maintained the administration, and use already established facilities (e.g. 140 traffic warden and other things). Thus the net-costs of the whole ordinance including salaries are approximately 2 million DKr. (Approximately 275,000 euros), hi practice the whole ordinance end up costing nothing to the Municipal, because there is a balance between the costs and the income from selling certificates and issuing fines. On the web page, www.citygods.dk, complete information about the City Goods ordinance can be found. It means that beside the printed matter people can get everything needed for applying for a certificate from the web pages. As the numbers show it's a popular feature, with 10,000 hits yearly.
THE CITY GOODS ORDINANCE IN THE FUTURE Status One major result of the City Goods Ordinance is that we found out how big a part of the transport work that's actually service and handicrafts transportation. This means, that the focus has turned somewhat from the outspringing "Loaded with care" to looking more at engine technologies. This means that only cars up to 3.5 tons and with engines younger than 7 years can drive freely in the medieval Copenhagen. Heavier cars can only drive if they utilize their capacity with a minimum of 60 % in average.
Remarks on the future A survey done among the Copenhagen citizens shows that they consider the heavy traffic as the most severe traffic problems in the capital right now. As the aim of the ordinance is to reduce the number and/or size and age of vans and lorries driving into the medieval Copenhagen, and thereby improve the city conditions (e.g. pollution, traffic-safety and congestion) it goes right into these problems. The traffic censuses show, that only half of the vans and lorries registered have a City Goods certificate. This indicates that the "City Goods ordinance" in it's present form only have a limited effect in controlling congestion, pollution, etc in city cores. However, seen on the total tonnage with City Goods certificates driven into the city, this tonnage has decreased by 1/3 in this final stage of City Goods (Table 5).
448
Logistics systems for sustainable cities Table 5: Decrease of the total number of cars with certificate and the total tonnage G reen + y ellow G reen + y ellow Green
Green
Numbers of cars
tola] tonnage
Green
Yellow
Average Numbers Dfcars
Yellow total tonnage
Yellow
at all
average
Average
Tonnage
Tonnage
1. Period
853
5,842,230
6849
2,983
13,860,940
4647
19,703,170
5,136
2. Period
942
6,801,055
7220
2,652
11,418,865
4306
18,219,920
5,070
3. Period
909
6,410,825
7053
2,016
6,397,164
3173
12,807,989
4,379
The reasons that only half of the cars in the traffic censuses are certified are due to different facts. The fairly cheap red certificate that allows any car to stop in the city core, some cars might drive without a certificate and cars only needs a certificate to stop. That means that any drive through is legal. We are working to overcome the difficulties with controlling and verifying the capacity utilization data given by the transporters, and thus make the whole ordinance less bureaucratic. As a consequence the city of Copenhagen might, once the City Goods has ended, start up a new voluntary pilot scheme trying to improve the element of capacity utilization with electronic means. A political decision to bring Environmental Zones into effect in Copenhagen by June 2004 combined with these elements from the City Goods Ordinance will make a strong and sufficient tool to regulate cargo traffic in the inner city.
Further information Results from the City Goods project will ultimo 2003 be published at the web page www.citygods.dk
33
THE E+ TRANSPORT ENVIRONMENTAL OPERATOR CLASSIFICATION SYSTEM
Kim Hassall, Associate Professor, Dept of Civil and Environmental Engineering Melbourne University; Director Urban Trucking Strategy, Australian Trucking Association, Australia
ABSTRACT As the land transport task, especially road is expected to increase significantly, by 2015, it would hardly be surprising that many policy makers will look to some form of urban transport environmental control long before this date. One method is via the introduction of a set of operational environmental hurdles. This is not a new concept. However, this paper proposes a new environmental initiative for urban transport operations. The scheme which is simple for both operators and regulators proposes an effective and meaningful operational rating system which reflects the efforts an operator may go to in their internal fleet environmental policies in purchasing equipment, daily workload planning, waste disposals policies etc. The three tiered environmental operator performance rating scheme, the E-plus scheme has three levels of operator segmentation. A good basic auditable level of environmental compliance, an E rating, an excellent rating E plus, and an exceptional rating would be an E double plus. Probably no fleets currently in Australia would earn a double plus rating at this time. Why should there be a road transport operator environmental rating scheme? Firstly as a measurable benchmark for the community and the fleet operators themselves. Secondly and a factor of growing importance is for the customers who are the buyers of freight services. Already customers are specifying in their tender requirements that transport operator environmental competencies and capabilities be listed. This may aid the selection of an operator for a specific task. For example; food sensitive freight handling may require specific food certification scheme adoption such as the HACCP classification system. Similarly an urban courier contract may specify, environmental credentials which sit under an ISO 14000 framework. However, what more specific operator differentiation criteria can be requested by
450
Logistics systems for sustainable cities
the customer? The E plus system is being designed for this very purpose. In Australia some regulators and teaching centres have begun to take an interest in this transport operator framework.
BACKGROUND Across many fronts, urban transport environmental concerns are being addressed through a range of local, State and federal regulations. However, a customer driven approach to the urban transport environment problem is possibly another avenue whereby road transport operator behaviour could be modified if it needed to be. Such a customer driven approach would need to be simplistic and effective for the freight buyers if it is to help deliver urban environmental policy outcomes, further and more successfully than the heavy hand of regulatory environmental policy intervention. This paper references developments of such a buyers' scheme in Australia over the last five years. What the Australian road transport regulators are doing? The National Road Transport Commission (NRTC) and the National Environment Protection Council (NEPC) jointly chair a government instrumentality called the Motor Vehicle Environment Committee (MVEC) (NRTC, 2003a). This committee nationally oversees the implementation of motor vehicle standards and environmental regulation within Australia. The rollout of emission standards are presented in Table 1 for all Australian States and Territories. Vehicle type
Table 1 Future Emission Standards 2002 to 2006, Introduction Dates Year 2002 2003 2004 2005
Diesel Fuel Light Trucks Medium and Heavy Trucks Petrol Fuel
Euro(II) Euro(III)
2006
Euro(II) Euro(III)
Euro(II) Euro(III)
Euro(II) Euro(III)
Euro(II) Euro(IV)
Euro(II)
Euro(II)
Euro(III)
Euro(III)
Standard
Limits on Emissions (g/Kwh) CO Hydrocarbons NOS Pre Euro(II) 4.5 to 20.8 1.1 to 3.8 6.8 to 8.0 Euro(H) 4.0 1.10 7.0 2.1 5.0 Euro(UI) 0.66 Euro(IV) 1.5 0.46 3.5 1.5 2.0 Euro(V) 0.46 Source: NRTC, personal communication., DoT, 1999.
Particulates 0.36 to 0.96 0.15 0.10 0.02 0.02
The E+ transport environmental operator classification system
451
The development of the national environment protection measure (Diesel) Further to the emissions standards being adopted by vehicle manufacturers, in-service emission performance, is also being addressed by legislation and regulation. One such in-service transport operator program is the National Environment Protection Measure (Diesel) (NRTC, 2003b) This program seeks to have any emission infringement fines going to a compulsory maintenance rectification program for that particular truck. The program has had its legislation drafted and is being enacted by the State authorities. National Heavy Vehicle Accreditation Scheme (NHVAS) This three part program was developed by the Road Transport Regulators in conjunction with the State Road authorities. The program is modular and comprises (NRTC, 2003c): 1. NHVAS Mass management (operational) 2. NHVAS Vehicle Maintenance (operational), and 3. NHVAS Fatigue management (In Draft) Certainly the second module has a major safety intent but it should also satisfy emissions compliance. These three regulatory programs are however, not integrated with International ISO quality Standards, that being either ISO 9001, ISO 14001 or ISO 2000. What has the Australia road transport industry been doing? Over the decade 1993 to 2002 the Road Transport industry association, initially the Road Transport Forum, now known as the Australian Trucking Association, (ATA) was also developing a national road transport accreditation program. This was called Trucksafe and it too had a maintenance compliance module, that was of a similar, and recognised standard to the NHVAS program. Further to this the ATA was developing a road transport maintenance workshop management scheme. This workshop accreditation model is in its pilot stage and currently only 3 operators, out of the 720 Trucksafe operators, enrolled in the Trucksafe program, have "Workshop Maintenance Accreditation". By the end of 2001 there were arguably 10 road transport certification schemes, two directly related to vehicle maintenance and none specifically designed to address the broader issues of a road transport Environmental Management System (EMS). hi 1995, Standards Australia attempted to integrate ISO 9001 with operational aspects of transport operations but this standard is in drastic need for overhaul especially since the NRTC's certification modules and the Trucksafe industry modules are fully operational. Hassall, Simpson and Barnesby 2001, attempted to show how this integration could be trialed. However, having a uniform, public domain standard, integrated under an International recognised framework is still some way off, perhaps more than five years away.
452
Logistics systems for sustainable cities
What do the transport customers want - is the tide changing? In late 2002 Australia was observing in advertised transport tender applications that there be non specific requirements for the specifications for environmental credentials by the transport operator. (The first sector to be seen to do this has been in the grocery related areas.) This is a commendable approach to require an "environmental capability statement" but also provides a novel way to differentiate between potential transport suppliers should selection difficulties arise where operators are offering similar customer services.
DEVELOPING A ROAD TRANSPORT FLEET ENVIRONMENTAL MANAGEMENT SYSTEM The first macro attempt to have an over arching, integrated EMS was through the Urban Fleet Scorecard (Appendix A). However, this approach although appealing to some operators and regulators, was not simplistic enough to capture the attention of that sector of the community that could really effect change, notably the large buyers of transport services, that is, the large customers. With this in mind the E+ Scheme was developed. It is still in its infancy, and has been loosely based on the Urban Fleet Scorecard approach.
What is the cost of an operator EMS? Many operators consider an EMS to be a cost imposition, however, this is an unfounded premise. Regular scheduled, and un-scheduled maintenance, usually adds to: • vehicle uptime, hence, better utilisation, • vehicle resale, • reduced emissions through the maintenance of injectors and oil filters, and • allows a marketing edge to particular customers. hi brief, whether or not a fleet environmental management system is set appropriately through an ISO 14001, an ISO 95 or a later ISO standard, or even a local national regulatory standard, a company 'should' find that financial savings can be attained through the adoption of a workshop certification/accreditation standard. The problem with an EMS focus alone is that vehicle maintenance covers a myriad of vehicle systems,eg • Suspension, • Braking, • Electrical, • Cooling, • Engine, • Exhaust and • Hydraulic Systems. Emissions are generally set in the domain of fuel injectors, oil and oil filters, fuel quality, exhaust systems and to a lesser extent tyre management strategies. This means that environmental considerations are only a subset of the total maintenance task.
The E+ transport environmental operator classification system
453
The development of E+ environmental management scheme 1. The development of many Australian de-facto road transport standards over the past decade has caused a very fertile environment upon which to build a generic EMS. Both the Industry Trucksafe and the National Road Transport Commission's NHVAS maintenance programs in particular, or an ISO certified maintenance workshop, can claim to have a basic level of compliant environmental maintenance. For this purpose this basic level for an EMS was set. It is described as an "E". 2. Further to the basic auditable maintenance and disposal standards, fuel quality and the use of low sulphur diesel, or even LPG, LNG or CNG, or even hybrid fuels may be considered nationally desirable and encouraged. This is considered to be one further increment in an operational EMS. The adoption of better emission engine design standards through an auditable purchasing program, based on optimal economic life, or full life cycle considerations, even just for engines, is adopted as at least a second hurdle within this second level of the E plus scheme. Therefore, considerations paid to fuel quality, low greenhouse emission fuels, or lower paniculate fuel and to engine emission standards generates a level of "E+" within the rating scheme. 3. Lastly, the management of fleet resources is considered to be the least implemented environmental management strategy. This "fit for purpose" consideration will determine the fleet capacity of vehicles and the mix of insourced versus outsourced, or contractor, operations. Usually to effectively determine this fleet mix and vehicle mix will be dependent on the scheduling tool. Many of these are commercially available: they range from Baan, 12, JD Edwards products to much cheaper systems. Very cheap and highly effective routing systems are available and have the ability to optimise the fleet mix for several hundred vehicles working from multiple depots. The adoption of such tools is generally marginal, often owing to the price for either the products or the consulting services, however, the benefits often range from some 11% improvement to 30% improvement when transport depot merging occurs. Average benefits for fleet kilometre and asset improvement often averages around 17% when no operational review has been conducted for some five or more years. The determination of fleet fit for optimised task is defined as "E++". Currently there is only one fleet in Australia that may reach the E++ standard but it may not fully comply with the full E or E+ standard. Within five years several dozen major fleets may gain such a certification level as E++. The 3 level of the E+ scheme are listed in Table 2. Currently research at the University of Melbourne is attempting to integrate ISO 14001 and the Standards Australia Road Transport Draft ISO standard. As well, a customer user survey will further define at a micro level, what sub-factors need to be incorporated into the E+ EMS. This work will be further reported to City Logistics at a future date.
Rating
Table 2 Segmentation levels of the E-plus rating system Operational Audit Achievement
E
Basics auditable maintenance, and waste disposals policies
E+
+Plus auditable purchasing strategy, and targeted adoption of low emission fuels
E++
+Plus adoption of software optimisation and task fit audit tools
Source: Hassall 2002
454
Logistics systems for sustainable cities
The intention of the E+ scheme is that an operational and environmental rating will be required by major customers in the quality capability statements of the operator when applying for the respective carrier contract. For major customers that may buy services from some 100 or more carrier providers, this EMS contract hurdle, is a very powerful hurdle in the sustainable operations of their road transport providers.
Ownership of the scheme All useful standards should belong in the "public domain" and they should not be proprietary standards. It is envisaged that the audit methodology will be further developed through a government grant system and the audit components of the scheme can be made available to all nationally registered quality auditors. However, an agency, a professional association, or a commercial registry may seek to hold and update the registry of applications to the standard as well as to list the certified transport companies that have been audited to the qualifying standards of the scheme. Very preliminary discussions have been held as yet to address this registry role, although regulatory interest in the scheme is somewhat more progressed. There may be differing timings of the audits for the transport companies to demonstrate their compliance to the E+ scheme. For example the E certification level may have an audit every three years, the E+ level an audit every two years, and the E++ level may require an annual audit. The determination of this frequency would be the outcome of future discussions with regulators and the environmental auditors.
CONCLUSION The development of a specific Road Transport EMS is beginning and is also long overdue. Australia, over the last 10 years has developed over 10 major road transport certification schemes. Rationalisation and simplification should occur as we currently have regulatory and operator based schemes. However, no scheme really fits that of a specific EMS and the proposed "E+ scheme" has been proposed and is gaining limited attention, despite its customer, and regulator appeal. The Transport customer is certainly a new player in the environmental transport debate but indeed a potentially very powerful player. The E+ scheme has tried to integrate existing maintenance and disposal policies at the basic E level. It requires a policy on clean fuel usage, and this does not heavily rely on any unstable technologies. Fleet purchasing policies, which often governments use to tightly regulate emission standards, forms the second rung of the E+ scheme. Lastly the E++ level, perhaps up to 10% of fleets could attain this rating, is predicated on an audit of fit for purpose and best fleet practice. Few national fleets within Australia would attain this rating in the next 12 months within Australia, and only a few internationally may do so at this time. The E+ scheme is the customer driven simplification of the Urban Scorecard proposal in Appendix A. Its simplicity is still being assessed amongst major buyers of transport services, although initial reactions are mixed amongst fleet operators and truck manufacturers. Customers and regulators are much more interested in the proposal.
The E+ transport environmental operator classification system
455
REFERENCES NRTC (2003a). http://www.nrtc.gov.au/publications/environ-operations.asp?lo=public Australian Livestock Transport Association (1998). Truck Care Management System, ALTA, Canberra. Australian Trucking Association (2002). Workshop Accreditation Auditor Matrix, Canberra (unpublished). Barnesby, J (1998). Road Transport Accreditation - 2000 and Beyond, Workshop Summary, Transport Workers Union, Melbourne (unpublished). Department of Transport Western Australia (1998). Fatigue Management for Commercial Vehicle Drivers, Operating Standards for Work and Rest in the Western Australian Road Transport Industry, Transport WA, Perth. Federal Office of Road Safety (1998). Independent Audit Framework for the Alternative Compliance Scheme, FORS, Canberra. Hassall, K.P. (1998). Accreditation Systems Need Streamlining, Australian Transport News, Vol 13, No 11, Publishing Services Australia, Brisbane. Hassall, K.P. (1997). The Quality Edge, Australian Transport News, Vol 13, No 1, September 1997, Publishing Services Australia, Brisbane. Hassall K.P., K. Simpson and J. Barnesby (2001). From Famine to Feast, A decade of Change in Australian Road Transport Certification, 5th International Quality & Innovation Conference, Melbourne. Hassall K.P. (2001) The Problem of Multi-Certifications Schemes in the Road Transport Industry: The STARS solution, 5th International Quality & Innovation Conference, Melbourne. Hassall K.P. (2002). Green Rating Incentives, Australian Transport News, Vol 17, No 10, Publishing Services Australia, Brisbane. NRTC (1993). Discussion Paper on Options for Improving Operator Performance, National Road Transport Commission, Melbourne. NRTC (1994). Alternative Compliance Workshop: Report to Participants, National Road Transport Commission, Melbourne. NRTC (1995). National Driver Licensing Scheme Evaluation, National Road Transport Commission, Melbourne. NRTC (1997). Increased Mass Limits: Compliance and Enforcement Issues, Discussion Paper, National Road Transport Commission, Melbourne. NRTC (1997). Alternative Compliance Policy Proposal, National Road Transport Commission, Melbourne. NRTC (1998). Proposed Mass Limits Second Hurdle - Evaluation of Mass management Compliance Assurance Schemes, National Road Transport Commission, Melbourne. NRTC (1998). Proposed Mass Limits Second Hurdle - Evaluation of Mass management Compliance Assurance Schemes, National Road Transport Commission, Melbourne. NRTC (2003b). http://www.nrtc.gov.au/publications/bulletin-37.asp?lo=public#M3 NRTC (2003c). http://www.nrtc.gov.au/publications/bulletin-40.asp?lo=public PACIA (1997). Responsible Care - Chemical Industry in Transition, PACIA, Melbourne. Queensland Department of Transport (1997). Fatigue Management Program - Pilot, Queensland Department of Transport, 1997. Road Transport Forum (1997). Trucksafe Accreditation Program, Small Fleet Operators Version, Road Transport Forum, Canberra. Road Transport Forum (1998). Greenhouse Challenge Workbook, Road Transport Forum, Canberra (unpublished).
456
Logistics systems for sustainable cities
Road Transport Forum (1999). Road Transport Forum Trucksafe, Mass Management Module, RTF, Canberra. Standards Australia (1995). Guide to AS/NZS ISO 9001:1994 for the Road Transport Industry, Standards Australia, Canberra. Standards Australia (1997). Integrating Quality and Environmental Management Systems, ISO 9001 and ISO 14001, Standards Australia, Sydney. Victorian Road Transport Association (1998). Hazard Analysis and Critical Control Point, Transport Management Australia, Victorian Road Transport Association, Melbourne. (unpublished) Victorian Road Transport Association (2000). TransCare Safety System - Introduction, Victorian Road Transport Association, Melbourne.
APPENDIX A: THE URBAN FLEET SCORECARD The concept of the "Urban Fleet Scorecard" was the forerunner to the E+ scheme. It was developed between the period 1996 to 2000 after undertaking several interviews with fleet operators, the Australian Road Transport Forum (now the Australian Trucking Association) and some Australian Regulatory authorities: one road transport authority and two environmental, authorities. The concept explored the assignment of a score (out of 100) for a fleets operation. This was an auditable fleet program score and the score was intended to be calculated each year. Although the Urban fleet scorecard is a preliminary step to the E+ scheme developments which are in progress to develop the scheme into a public domain piece of software within Australia by a major fleet operator. The scheme comprised some 19 factors and various elements do contain sub-factors. The list was developed as those fleet factors that were considered important by: • Transport Managers • Transport maintenance workshop managers • Environmental regulators, • Road transport regulators • Quality Auditors, and • Occupational Health and Safety Consultants. Calculating a score was based on a set of surveyed allocated class weights derived from three observers' perspectives: 1. A fleet transport perspective 2. An Occupational and Safety perspective, and 3. An environmental perspective. There are large variances with the weights used across each of these segments although the fleet, health and environmental weights all add to 100. Item 13 attracts a very large score as the first (non-listed) sub-division is fatal accidents, and this gains a score of 35% on the health set of class weights.
The E+ transport environmental operator classification system
457
Table A: The Construction of the Urban Fleet Scorecard Elements of the Urban Fleet Scorecard Average of 3 Class Weights 1. Total Fleet Fuel Consumption 19.3% 2. Fleet Fuel Consumption Rates 11.3% 3. Aerodeflector adoption 5.0% 4. Leaded Petrol Use 3.8% 5. Vehicle Oil disposal policy 1.8% 6. Vehicle Battery disposals policy 0.9% 7. Other parts disposals 1.0% 8. Tyre retread policy in place trucks and prime movers 3.0% 9. Tyre retread policy in place trailing equipment 3.5% 10. Auditable maintenance programs 7.7% 11. Alternative or low emission fuel trials in place 1.9% 12. Involvement in Greenhouse or abatement programs 2.3% 13. Accidents recording and corporate policy in place 22.9% 14. Vehicle smoke infringements recording system 3.3% 15. Vehicle noise infringements recording system 3.2% 16. Vehicle policy for washing/cleaning at specialised 2.3% facilities 17. Vehicle maintenance depot yard sealed from leakage 4.3% 18. Company policy on Tollway use 1.9% 19. Clean company vehicle livery policy in operation 0.6% Source: Hassall 1997 In 2001, the road transport fleet rated by the national Australian Fleet Managers' Association for that year would have rated some 70 to 75 points on the Urban Fleet Scorecard.
This page is intentionally left blank
34
RELATIONSHIPS BETWEEN GOODS DISTRIBUTION AND PUBLIC TRANSPORT IN URBAN AREAS - THE CASE OF A HYPERMARKET IN PORTO
Alvaro Costa, Faculdade de Engenharia da Universidade do Porto, Portugal Sandra Melo, Faculdade de Engenharia da Universidade do Porto, Portugal
ABSTRACT The behaviour of the key stakeholders in city logistics, as defined in Taniguchi et al. (2001) (e.g. shippers, residents, freight carriers and administrators), can be affected by the decisions taken by other stakeholders. The case presented in this paper reports on the result of a bus line extension up to the main entry of hypermarket in Porto Metropolitan Area. Passengers reported changes in their shopping habits and senior people reported the utilisation of the home delivery service for the first time after the line extension. It appears that the existence of home delivery service provided by the hypermarket contributes to the increase of the patronage in public transport.
INTRODUCTION Between 1996 and 2000, the average income increased 1.2% in Portugal (Tavares, 2002) and as a consequence the number of vehicles increased 6.7%, achieving 470 vehicles per 1000 inhabitants. In Porto Metropolitan Area (AMP) the rate of car ownership in 2000 was 403 vehicles per 1000 inhabitants with 368 in the city of Porto (INE, 2000). This great increase in the usage of private cars for the daily trips has negative consequences in terms of urban sprawl. Furthermore, between 1991 and 2001, the utilisation of public transport in AMP decreased from 42% to 28% as the individual transport increased from 31% to 52%.
460
Logistics systems for sustainable cities
In an effort to change this tendency, the Sociedade de Transportes Colectivos do Porto (STCP), the bus transport operator in Porto, conducted several market studies and concluded that many residents were prepared to use the public transport for most activities, but that they need to use the private car in some links of the chain of their daily movements. The need to use the car in one link of the chain of daily movements seems to determine the utilisation of the car in all other links of the chain. Other issues which determine the choice of the transport mode like the reliability of the public transport chain (Rietvield et al, 2001) are not considered in this paper. The activities that usually determine the utilisation of the private car are needs such as taking children to school and shopping activities. Trips related to these activities appear to contribute more to the traffic congestion than any others (Rietvield et al 2001). This was felt in the bus operations during school holidays and during peak shopping periods. The objective of this paper is to show how the completion of missing links between the circuits of the delivery of goods and the mobility of residents using public transport (by overlapping goods and passengers networks) can induce people to change their behaviour and opt for public transport. The case presented here reports on an experience of a bus line extension to a shopping centre. Among other consequences of the extension, such as an increase in patronage, senior people reported changes in their shopping habits explained by the existence of home delivery and public transport services.
PORTO METROPOLITAN AREA The Porto Metropolitan Area is located in the northern seacoast of Portugal and it is formed by 9 municipalities, Espinho, Gondomar, Maia, Matosinhos, Porto, Povoa de Varzim, Valongo, Vila do Conde and Vila Nova de Gaia. The Metropolitan Area occupies a total of 817 km2 of which 40% is mainly urban. Between 1991 and 2001, the population increased 7.6%, from 1,168,000 inhabitants to 1,257,000 inhabitants (INE, 2002). The city of Porto is the heart of this dynamic metropolis and has an area of around 42 km2. During the last decade it has decreased its population at a rate of 13.1%, decreasing from 302000 inhabitants to 263,000 (INE, 2002). At the same time the average size of families decreased from 3.7 persons in 1981, to 3.3 in 1991 and 2.9 in 2001, which meant that, in spite of the diminishing of population, new dwellings were still needed to provide for the growing number of families.
Relationships between goods distribution and public transport in urban areas
461
Figure 1 Geographical location of AMP
URBAN SUPPLY IN PORT METROPOLITAN AREA Shop owners in the inner city of Porto are going through serious difficulties mainly because of the lack of accessibility to the town centre and because other shopping options have appeared, such as shopping centres in the outskirts of the city, that are now available. Porto is an old city with a very rich historical centre where the streets are very narrow and the distance between intersections is very short, making it difficult to accommodate urban supply. The insufficient width and the inadequate layout of the streets, together with a deficient transit regulation and the lack of proper law enforcement, produces a proliferation of transgressions, mainly related to double lane parking in streets by freight vehicles, which creates great problems of traffic congestion. The areas reserved for the loading and unloading activities are not adapted to the present needs of freight vehicles. The root of the congestion problems, relating to loading and unloading operations of freight vehicles, lies not only will the physical characteristics of the road network and on law enforcement, but also on the poor organisation of the distribution of goods. The utilisation of information systems is scarce and ordering the goods often requires the physical presence of suppliers (Melo, 2003). Some commercial sectors such as the pharmacies are exceptions, as they developed partnerships to organise a more effective delivery system. On the other hand, shopping centres located in the periphery of the city of Porto are very successful, because they have better accessibility, a good offer of parking facilities and have longer opening hours. The competition between shopping centres is fierce and the conditions of access to their commercial areas and parking facilities affect the choice of the clients. Most shopping centres are located in the outskirts of the city of Porto (Figure 2), but already in other municipalities, as
462 Logistics systems for sustainable cities the authorities in Porto do not give permission for the construction of this kind of commercial facilities within the city limits.
Figure 2 Geographical location of the Hypermarkets in Porto Metropolitan Area In AMP, the hypermarkets generally anchor the shopping centres. There are five large hypermarkets in the AMP, three of which are located in the south bank of the River Douro, in the municipality of Vila Nova de Gaia — Carrefour, Jumbo and Continente. Jumbo and Continente are inside two shopping centres, ArrabidaShopping and GaiaShopping, respectively. The remaining hypermarkets of AMP, on the northern bank of the River Douro, are also located within the facilities of shopping centres, Jumbo in MaiaShopping and Continente in NorteShopping. The closest hypermarket to downtown Porto is located in the border with the municipality of Matosinhos, Continente - NorteShopping. This hypermarket just like the ones already mentioned, with the exception of Carrefour, is located in the buildings of a shopping centre, out of the city limits and quick accessibility is assured by its close proximity to the main roads of AMP. In an attempt to increase the number of customers, hypermarkets are trying to develop an additional range of services. The present strategic location of the hypermarkets combined with new shopping solutions, such as the development of e-commerce and the home delivery of goods (Gould, 1998), demonstrates that they will continue, to compete aggressively with other hypermarkets in the region.
Relationships between goods distribution and public transport in urban areas
463
THE CASE STUDY The case presented here reports on the result of a bus line extension up to the main entry of NorteShopping, located in the municipality of Matosinhos. The transport operator STCP is a public transport company owned by the central government of Portugal and operates in a regime of monopoly of the bus operations in the city of Porto. The enterprise also provides some services in the outskirts of the city competing with private operators. The transport network covers an area that extends 10 km north and 6 km south of the city, including the neighbouring municipalities of Matosinhos, Maia, Valongo, Gondomar and Vila Nova de Gaia. STCP operates a fleet of approximately 600 buses, produces around 35 million vehicle kms per year and has around 2,600 bus stops. In recent years we witnessed an improvement in the performance of the enterprise, both in terms of efficiency of production and effectiveness of the service, and also in the number of total staff which decreased from the 3,672 reported in 1990 to 2952 in 1996 and 2,240 in 2002. At the same time the number of drivers increased from 1,229 in 1996 to 1,334 in 2002 and represent now 60% of the workforce comparing to 42% in 1996. The production has been around 30 million vehicle km per year but the increase in the number of drivers made it possible to cut extra working hours, as in 1996 around 10% of the production was made by staff working overtime. The fleet was modernised by the acquisition of buses in operational leasing that reduced the average age of buses from 13.6 years at the end of 1997 to 6.2 years in December 2002. All of this could be accomplished by a strong investment in IT systems in order to control the communications and acquisition of new facilities to centralise the administrative offices. Information systems will soon improve with the current installation of GPS in fleet operations. The enterprise is very much oriented to customer service, the corporate image changed and now the social recognition is high. Although the patronage is still decreasing, it changed the trend that began in the 90's and now it has stabilised.
Continente-Norte Shopping The hypermarket Continente opened in 1985 and it has a sale area of 10,800 m 2 . Before the construction of NorteShopping, this hypermarket had an exclusive function of anchor, where many residents in AMP went to make their daily purchases. With the opening of the shopping centre, on 21st October 1998, the hypermarket has been sharing the anchor function and that contributes to the traffic jam in that area provoked by private vehicles that try to access the hypermarket or to the shopping centre or both. The shopping centre has an area of 75,000 m2 occupied by 288 stores, an entertainment centre and 5000 parking places not considering the hypermarket.
464
Logistics systems for sustainable cities
The extension of line 36 The bus transport operator (STCP) launched in September 1996 a bus line (number 36) linking residential areas to a main shopping area at one end of the line, but the patronage was very low. In 1998, in an effort to improve the effectiveness of the operation, a functional analysis of the activities in the streets crossed by the line was carried out. The line started in a commercial area in the inner city and crossed a very dense area of academic facilities and residential houses. The patronage was low because it took too long to go from the residential areas to the commercial area.
Figure 3 Line 36 of STCP before and after the extension So it was decided to extend the line 1200m towards the NorteShopping, outside the city limits (Figure 3). The line was extended on 19th of October 1998 and now it ends at the shopping centre.
Relationships between goods distribution and public transport in urban areas
465
Figure 4 Monthly data on the number of ticket validations Increasing the patronage of a bus line takes time to develop. In this case, since the first day of the extension of the line, there was an increase in the number of passengers. Figure 4 shows the number of ticket validations during a period of 13 months. The number of validations increased more than 30% between February 1998 and February 1999. Travel cards are not included in the figures as they are not validated. Goods distribution and mobility The increase of public transport utilisation in line 36 corresponds to a change in the habits of the new clients, especially those who live near the transport route. It is difficult to identify the most important factor contributing to the rise in the number of passengers. Many explanations can be found because there was no previous concern to evaluate this experience considering the behaviour of the passengers. However, passengers reported changes in their shopping habits and senior people reported the utilisation of the home delivery service provided by the hypermarket. Line 36 crosses a residential area of very expensive detached houses inhabited by elderly people relying on relatives for their shopping activities, mainly during weekends. The extension of the bus line to the shopping centre, complemented with the home delivery service provided by the hypermarket, rebuilt a sense of independence to people previously relying almost exclusively on relatives to solve their shopping needs. Instead of going shopping during the weekends in a family car, some residents met their shopping needs by bus and used the home delivery service at a cost of 5 euros. Home delivery is a quite recent service provided by Portuguese hypermarkets. They developed home delivery services and E-commerce options for their clients and this was especially important for those with less mobility.
466
Logistics systems for sustainable cities
Figure 5 Logistical Organisation Schemes Everyone has a chain of movements which corresponds to their personal daily activities. This chain is different everyday of the week and the utilisation of private car provides the flexible means for meeting those daily needs. Sometimes, because of one of the car dependent activities in that chain of movements, people resort to the use of car transport for the remainder activities. This fact is widely recognised, but what is curious about this example is that one of the movements (i.e. shopping), that often determines the utilisation of the car, can be solved through public transport when complemented by home delivery services. Figure 5 illustrates what we have in mind describing the options available for the movements to the hypermarkets before and after the extension of the bus line. Before the bus line extension, customers had three options: to go and to return by private car or taxi with their goods, or order them through the internet. After the extension, there was one more option that consisted of using bus transport together with the use of home delivery services offered by the local hypermarkets. We do not have enough data to measure the impact of home delivery service on the utilisation of public transport. Although, it appears that the existence of home delivery contributes to the increase of the patronage in public transport. Customers reported the utilisation of this service for the first time after the line extension.1
CONCLUSION The behaviour of the key stakeholders in city logistics, as defined in Taniguchi et al. (2001) (e.g. shippers, residents, freight carriers and administrators), can be affected by the decisions taken by other stakeholders. First, as reported in the case of AMP, this relates to a change in the behaviour of residents induced by the decision of a public transport operator. Secondly, the services of home delivery of goods, already in place, also contributed to this result. Clients were not the only ones showing satisfaction with the line extension. Also their relatives, now not needing to go shopping during weekends, reported the importance of the extension.
Relationships between goods distribution and public transport in urban areas
467
The bus line was extended with the objective of improving the profitability of a transport operator but, in the end, it resulted in clients reporting a change in their shopping habits as they need not rely on other people to go shopping. Not having to rely on relatives to go shopping during the weekends, because other solutions were implemented by the bus transport operator and the home delivery services of the hypermarkets in place, provides a sense of independence to senior people and increases the utilisation of public transport. Also the goods can be home delivered in vans or small lorries instead of being carried by private cars. Many hypermarkets created a service of home delivery in order to enhance E-commerce opportunities and provide new services to other type of customers. This service is a complement not only to the E-commerce, but also to the clients who use the public transport the missing link in the intersection of the two networks can be, as reported in this case, for other customers, the availability of a public transport service. Independently of the dimension and impact of this extension of this particular bus line in terms of present or future shopping habits, which we could not measure, some clients reported the using the home delivery service for the first time. The extension of the line was decided without any previous collaboration between urban stakeholders. However, much of the traffic congestion and many environmental problems in urban areas can be solved by overlapping the networks in order to identify the missing links between the circuits of the chain of goods distribution and the mobility of residents provided by the public transport networks. Finally, we would like to point out that it is important to measure all the consequences of the decisions taken by different urban stakeholders in every aspect of urban life. Unfortunately, very often, as in the case reported, great concern was taken to evaluating the result in terms of number of passengers but little attention was given to the reasons behind the results achieved (Woudsma, 2001)
REFERENCES Gould, J. (1998). Driven to Shop? Role of transportation in future home shopping, Transportation Research Record, 149 - 156. INE (2000). Inquerito a Mobilidade da Populagao Residente 2000, Instituto National de Estatfstica, Portugal. INE (2002). Recenseamento Geral da Populagao e da Habitagao 2001, Instituto National de Estatistica, Portugal. Melo, S. (2003). XJma logistica colaborativa para a cidade, MSc Thesis, Faculdade de Engenharia da Universidade do Porto (unpublished). Rietvield, P., F. R. Bruinsma and D. J. van Vuuren (2001). Coping with unreliability in public transport chains: A case study for Netherlands, Transportation Research Part A, 35, 539 -559. Taniguchi, E., R. G. Thompson, T. Yamada and R. van Duin (2001). City Logistics - Network Modelling and Intelligent Transport Systems, Elsevier, Oxford. Tavares, C. (2002). IV Congresso dos Empresdrios do Centro, Conselho Empresarial do Centre, Ministerio da Economia, Portugal. Woudsma, C. (2001). Understanding the Movement of Goods, Not People: Issues, Evidence and Potential, Urban Studies, 38 (13), 2439 - 2455.
This page is intentionally left blank
