19 th European Symposium on Computer Aided Process Engineering – ESCAPE19 J. Jeowski and J. Thullie (Editors) © 2009 Elsevier B.V./Ltd. All rights reserved. The supply-chain pick-up and delivery problem with transshipments Rodolfo Dondo a , Carlos A. Méndez a , Jaime Cerdá a* a INTEC (UNL-CONICET), Güemes 3450, 2500- Santa Fe, Argentina, * Email: jcerda@intec.unl.edu.ar Abstract In this work, a strict MILP formulation for the supply-chain pick-up and delivery problem with transshipments is presented (SC-PDP-T). This problem adds the option for transferring goods from one vehicle to another. This additional flexibility is very attractive for the optimal management of supply chains. Based on the novel features of the SC-PDP-T, this work addresses the major modeling and solution issues related to this problem by presenting a new MILP-based strategy to find the set of decisions to optimally manage complex multi-site distribution systems of moderate-size. Keywords: Supply-chain, pick-up and delivery, transshipments. 1. Introduction The cost-effective management of multisite production system is a complex task that needs to be aided by efficient computational tools. In this direction, the pickup and delivery problem (PDP) has been one of the most studied network logistic problems in the transportation-research literature [1, 2]. In the emerging area of enterprise-wide optimization, a great deal of effort is focussed on the optimization of complex supply chains. So, in the last years, the inherent features of real-world supply chains have motivated the development of numerous and more realistic variations of the classical PDP. Following this trend, the so-called supply-chain PDP (SC-PDP) have been defined by [3] aiming at generalizing the typical PDP by considering alternative supply sites, inventory constraints, multiple visits to the same site and multiple commodities. Sometimes, forcing each load to be directly transported from its source to its final destination by using a single vehicle is a hard assumption. This limitation explains why the PDP, a widely studied problem in the transportation research area, has not been widely used in supply chains applications. The possibility for goods to be transferred from one vehicle to another adds a higher flexibility to the operation that could improve the overall productivity by exploiting the interaction between vehicles at specific transfer points. Obviously, if the added flexibility reduces the total load-transported and travel times, one could try to find favorable conditions for implementing a support system that allows transfers instead of rigid PDP formulations. So, the PDP is here extended to consider transfer points, i.e. distribution centers, where some vehicles can drop its load to allow others to pick up it later, as defined by [4]. The modeling of this new feature generates a new logistic problem called the SC-PDP with transshipment (SC-PDP-T). 1009