A Threshold Inventory Rationing Policy for Service - Differentiated Demand Classes

Motivated by a study of the logistics systems used to manage consumable service parts for the U.S. military, we consider a static threshold-based rationing policy that is useful when pooling inventory across two demand classes characterized by different arrival rates and shortage (stockout and delay) costs. The scheme operates as a (Q, r) policy with the following feature. Demands from both classes are filled on a first-comefirst-serve basis as long as on-hand inventory lies above a threshold level K. Once on-hand inventory falls below this level, low priority (i.e., low shortage cost) demand is backordered while high priority demand continues to be filled. We analyze this static policy first under the assumption that backorders are filled according to a special threshold clearing mechanism. Structural results for the key performance measures are established to enable an efficient solution algorithm for computing stock control and rationing parameters (i.e., Q, r, and K). Numerical results confirm that the solution under this special threshold clearing mechanism closely approximates that of the priority clearing policy. We next highlight conditions where our policy offers significant savings over traditional ‘round-up’ and ‘separate stock’ policies encountered in the military and elsewhere. Finally, we develop a lower bound on the cost of the optimal rationing policy. Numerical results show that the performance gap between our static threshold policy and the optimal policy is small in environments typical of the military and high technology industries. This research was supported in part by the U.S. Navy under Contract # NOV391-96-M-M04, NSF CAREER Award #9602072, NSF grant #0075391, and the Fishman-Davidson Center for Service Operations Management at the Wharton School. Special thanks to the following people from various organizations within the military Sandy Leggieri, Gary Burchill, Jere Engelman, and Mike Puoy. The authors also acknowledge the detailed and insightful comments of Bill Lovejoy, the AE and two anonymous referees on earlier versions of this paper.