Modeling e-Logistics for urban B2C in Europe Francesc Robusté, Dante Galván and Andrés López-Pita CENIT (Center for Transportation Innovation) Universitat Politecnica de Catalunya (Technical University of Catalonia) Civil Engineering School Jordi Girona 1-3, edifici B1 – 08034 Barcelona (Spain) Tel.: + 34 93 4017104, Fax: +34 93 4015675, <f.robuste@upc.es> Abstract. The paper quantitatively models e-logistics supply chains for urban business-to-consumer relationships in the case of grocery product delivery. Orders are placed through internet, but deliveries are made at the customer’s home. The model takes special account of the effect produced on logistics variables (distances, times, costs, shipments, stops, etc.), by the way in which orders are received, and includes the delivery in time windows in urban areas. The findings are of interest both to transportation planners and e-logistics industry. Planners should design parking space and loading/unloading procedures so that the deliveries are made efficiently and with little traffic perturbation, and the industry should start getting ready to understand the high cost of this quality of service: the “picking” process has a high cost that cannot be assumed in high volumes and everyday the good (the street is contemplated as a scarce social good) and bad (double parking, bus lane blocking, etc. causing additional traffic congestion) use of the streets will likely cost more money. An application to Barcelona, Spain, for the food distribution sector (supermarkets) shows that, for the previously stated reasons, the current industry procedures are only feasible for very low demand share (less than 10%). 1. INTRODUCTION E-commerce is presented as a source of demand for road or street space, characterized by practices such as just- in-time, zero stock, personalized delivery with limited windows of time, etc., which results in an increased need for transport and pick-up-and-delivery (Robusté, 2002). When a large number of half-empty delivery trucks is added to the circulation in an urban environment, the local authorities should question, or at least evaluate, the feasibility of the model of service currently offered by virtual shopping. In some sectors, such as consumer goods, the arrival of dot.com competitors (initiated on the Internet) has caused a massive incursion of e-traditional companies (initiated in physical stores) which sell their products as part of a strategy to gain positioning online. The special characteristics of the order (volume, picking, manipulation, etc.) and of the service (hours open, personal delivery, etc.) make this sales channel a true challenge of logistics for the e-companies. In this scenario, the decisions regarding logistics have important financial and economic repercussions not only on the supply chain, but also in the entire business environment and often determine the success or failure of the business. The low level of existing demand and the need to stand out in customer service, added to the critical nature of the decisions, make it indispensable to have tools for determining the resources needed and the costs associated with each level of service. The e-company cannot adjust its level of service by trial and error, because the online client is only one click away from shopping at the competition. 2. MODEL The model uses the method of continuous approximations (Daganzo, 1999) and follows the steps of a classic process which determines, among other things, the main indicators of the system’s demand (volume and quantity of orders, distribution in time and space, etc.) and supply (number and type of vehicles, rate of service, etc.), to finally make an analysis based on queuing theory. This method presents the density of points to visit as the main demand descriptor, hence it is necessary to know the attraction of online demands and the surface area represented by each of the zones within the defined area of study. The area of study is divided into n delivery zones (subscript i), which are determined according to the functional, geographical and economic characteristics, etc. TRB 2003 Annual Meeting CD-ROM Paper revised from original submittal.