Models for managing product upgrade in closed loop supply chains

In this dissertation, we develop strategic and operational level decision making tools to help manage product upgrades in closed loop supply chains under limited availability of upgrade resources. The models developed in this research are mainly applicable to products that are modular, repairable and expensive. This research was directly motivated by a research project done for the United States Coast Guard (USCG) to help them plan an aircraft upgrade from an older to newer version. The upgrade also involved upgrading an aircraft component (gearbox) in order to achieve higher operational efficiency. Apart from developing the planning tools, we also study how the decisions made by these planning tools vary under different levels of operational flexibility. We first develop strategic level decision making tools that can be used to plan the rate of product upgrades in closed loop supply chains under limited availability of upgrade resources. We show that in closed loop supply chains, the optimal rate of product upgrade depends on current availability of upgrade resources and future supply of spare parts and that the flexibility available to operate in different product-part configurations during product upgrade can compensate for scarcity of upgrade resources. We then develop models of operational level decisions during product upgrade over a shorter planning horizon, under different levels of operational flexibility. We first develop a model for the discrete version of the problem to study the exact allocation of upgraded and regular components to products when they fail. Here, we develop the optimal product and component hold back policy for a given level of operational flexibility and show that a myopic policy of always keeping products operational or using up all the upgraded components immediately may not always be optimal. We then study a continuous time version of the problem and provide numerical results. Finally, we consider a maintenance network of repairable parts with age-dependent maintenance times, motivated by a maintenance network at the USCG. Studying the impact of inventory of repairable parts (say gearbox) on system performance shows that as the inventory of gearboxes increases, the aircraft availability may vary non-monotonically, which is counterintuitive. Though we have not been able to find any analytical results to prove the same, we present numerical evidence showing this non-intuitive result. We leave this problem as an open question for future research. In summary, this research develops decision making tools to assist the optimal management of product upgrades in a closed loop supply chain. Managerial insights are developed using data motivated by the US Coast Guard problem context