A simulated annealing methodology to distribution network design and management

This paper describes the PLOT (Production, Logistics, Outbound, Transportation) design system. The system addresses a class of distribution network design problems, which is characterized by multiple product families, a central manufacturing plant site, multiple distribution center and cross-docking sites, and retail outlets (customer zones) which demand multiple units of several commodities. The resulting system focuses on two key stages: the planning stage where we incorporate a strategic based decision-making process of selecting the "best" set of distribution centers and cross-docks to operate. The second phase consists of the execution stage that includes an operational based decision-making process. In this phase, the model decides the required quantity of product families that need to be transported from the plant to distribution centers and transshipped to cross-docks from warehouses, and later distributed to customer outlets. The distribution system design considered here is derived from current operations of a major retailing organization that manages products for nationwide distribution. The PLOT system developed to implement the model provides for a high degree of user interaction in the generation of solutions. The overall system generates globally feasible, near optimal distribution system design and utilization strategies utilizing the simulated annealing (SA) methodology. This study makes two important contributions to the SA literature, First, we extend the breadth of applications by studying a new combinatorial problem that incorporates cross-docking in a supply chain environment. Second, we systematically evaluate the computational performance under a variety of problem scenarios and SA control parameter settings. (C) 2002 Elsevier Science B.V. All rights reserved.