The industrial research project RESPET aims to develop quantitative approaches for a central problem in freight transportation : the management of door to door freight transportation services. The problem addressed is characterized by a global optimization of the transportation chains (and not segment by segment) and by the existence of simultaneous multiple objectives. The decision-aid methods proposed will be validated on scenarios defined with our industrial partners. RESPET is subject to a request for labelling with the world-class competitive cluster i-Trans.
Why dealing with optimization problems in freight transportation? Freight transportation plays a central role in social and economical activities. The tonnage of goods transported has increased steadily during the last twenty years. But transportation also leads to many nuisances in terms of safety, congestion and the environment. Therefore increasing the transportation system performance at a given level of resources is a strategic issue.
Why consider a multi-objective approaches ? The management of a transportation system is a difficult task which necessarily involves a multi-facet analysis. For the analysis of a transportation system, this concern is clearly addressed using mainly qualitative methods. However, despite numerous works on quantitative methods, only one objective is considered in most optimization approaches. In other words a single function is optimized.
In this project we take into account the simultaneous conflicting objectives related to economical and environmental aspects but also to the service quality. Specifically, we assume that a consortium of logistics operators manage a long distance transportation network (LDTN), typically a railway or a sea transportation network linking different multimodal terminals. The LDTN characteristics are known a priori. The consortium aims to develop door-to-door transportation services using the LDTN. Given a set of demand, transportation chains must be determined from their pick-up location to their destination. These chains include pre-or post routes rooted at the LDTN terminals and starting from/ending at their pick-up/delivery points. These chains, possibly multimodal, may be more or less complex and are not known a priori. They may require the use of a multi-level transportation system with intermediate cross-docking terminals. Considering different objectives without any a priori ranking, we will aim to develop models and effective decision support methods and to evaluate them. Multi-objective problems will be considered from two points of view. First we will consider the case where the consortium is subject to regulation. Second we will assume that the consortium acts as a leader, but that customers have some choice between different delivery systems. In the first case we will propose multi-objective optimization models and approaches, whereas in the latter we will consider a bilevel optimization approach. Regulated and unregulated cases could be combined when there is collusion between the decision agents. These two approaches will provide decision makers with more relevant solutions to their own problems. Decision support methods will be validated on three scenarios in which the multi-modal services will be managed by using only the resources of the consortium (basic case), by sub-contracting some logistic operations or by unsing a pooling strategy between the consortium and their partners.
Madame Luce Brotcorne (INSTITUT NATIONAL DE RECHERCHE EN INFORMATIQUE ET EN AUTOMATIQUE - (INRIA Centre Lille-Nord Europe)) – email@example.com
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
DHL SUPPLY CHAIN DHL SERVICE CENTRAL
INRIA INSTITUT NATIONAL DE RECHERCHE EN INFORMATIQUE ET EN AUTOMATIQUE - (INRIA Centre Lille-Nord Europe)
LAAS CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE - DELEGATION REGIONALE MIDI-PYRENEES
LIA UNIVERSITE D'AVIGNON ET DES PAYS DE VAUCLUSE
Help of the ANR 559,803 euros
Beginning and duration of the scientific project: January 2012 - 36 Months