This thesis is concerned with coordination aspects of supply chain management and, in particular, explores lateral coordination across the supply chain. We propose a new framework for supply chain coordination, including mechanisms and policies that leverage finite capacity scheduling to provide more effective coordination. These mechanisms are studied under a number of different supply chain configurations, supplier-customer relationships, load conditions, and degrees of uncertainty.
The proposed coordination framework is designed for decentralized systems of self-interested and rational entities. It is based on agent technology, where several software agents, each responsible for a particular supply chain entity (e.g., a shop or an entire plant), cooperate and coordinate to maintain consistent schedules. Key advantages of this approach are its assumption of decentralized control (i.e., each supply chain entity, modeled as a software agent, is an autonomous entity), the possibility of concurrent or asynchronous execution, the flexibility of control mechanisms, and the ability to reconfigure and extend the supply chain model.
We have tested and compared the performance of our coordination policies according to a number of indicators, such as profit (sales revenue minus costs), leadtimes, customer satisfaction, and the ability to accurately forecast order completion. Our empirical experiments indicate, with high levels of statistical significance, that policies which synchronize finite capacity schedules across the supply chain can reduce the number of tardy orders by up to 50 percent, cut leadtimes by up to 30 percent, and provide a significant increase in profit over traditional leadtime-based coordination approaches.