Group for Research in Decision Analysis

Integrated Production and Distribution Scheduling in a Supply Chain

Zhi-Long Chen

There is a large body of literature on integrated production-distribution models. However, most of the existing results deal with strategic or tactical level decisions, and few results exist for integrated problems at a detailed order-by-order scheduling level. On the other hand, in many applications involving make-to-order or time-sensitive products, the production and assembly of final products and the delivery of finished products to customers are very closely linked and need to be considered jointly at detailed scheduling level in order to minimize total production and distribution cost while maximizing delivery lead time performance. Application examples include computer assembly and delivery, food catering services, newspaper printing and delivery, mail processing and delivery, and production and delivery of some perishable industry chemicals. In this talk, we will present two integrated production-distribution scheduling models that we have studied recently. The first model involves a single supplier-single customer supply chain where the supplier needs to process a set of orders received at the beginning of a planning horizon and deliver the finished orders to the customer. The objective is to minimize the total transportation cost while meeting the given delivery deadlines of the orders. The second model involves a supply chain with multiple overseas plants and a domestic DC where the decisions include assigning the orders to the plants, scheduling the processing of the orders at each plant, and scheduling finished order delivery from plants to the DC. The objective is to optimize the tradeoff between the total production and distribution cost and a delivery lead time performance measure. We will talk about solvability of various cases of the models and provide heuristic algorithms for some intractable problems involved. Worst-case or/and asymptotic performance of the heuristics will be analyzed.