Optimizing the supply chain configuration and production-sales policies for new products over multiple planning horizons

The objective of this paper is to jointly optimize supplier selection, pipeline and safety stock inventories, and production-sales policies for new products over multiple planning horizons such that the total profit over the product's life-cycle is maximized. A mixed-integer nonlinear programming model is proposed that not only allows for adjusting the level of pipeline and safety stocks at different supply chain stages, but also for switching suppliers and/or modifying their demand allocations in different planning horizons in response to changes in the new product's demand. A representative supply chain network for computer assembly is then used to illustrate the applicability of the model. The findings from our numerical experiments with the model can potentially have important implications for future research and practice. In contrast to the traditional understanding of flexibility, the results indicate that the total profit may not necessarily be an increasing function of the number of planning horizons. The performance of commonly used heuristic policies for supplier selection as well as different myopic and build-up production-sales policies are also compared with the model's prescribed solutions in order to provide insight on when such heuristic policies and combinations thereof can be effective. The interdependency between supplier switching cost and the number of build-up periods is also illustrated.