Repairable inventory theory: Models and applications

Repairable inventory theory involves designing inventory systems for items which are repaired and returned to use rather than discarded. Such systems are composed of items which are typically less expensive to repair than to replace, and are considerably more complicated than traditional consumables inventory systems. The typical problem is concerned with the optimal stocking of the repairable parts and the location of these stocks, given that there may be multiple locations. An added dimension to the problem is the determination of the size and location(s) of the repair capacity for these parts. Further, different performance measures may be used, such as cost, backorders, and availability. There are many complicating factors in the design of repairable inventory systems, for example, not all failed units can be repaired and put back into service, some will be condemned and have to be replaced by new procurements. Various solution approaches have been developed to solve the problem, few have been implemented in practice, and no single model has addressed all or most of the complicating factors. Recent trends in the repairable inventory environment, environmental trends and regulations, and trends in product design are calling some of the assumptions of earlier models into question. In this paper we discuss the existing body of literature on repairable inventory, examine the various models proposed and the major assumptions made in those models, and classify them according to their solution methodology, single versus multi-echelon, and exact versus approximate solutions. It is intended to aid practitioners and researches in identifying the sources for existing methods and the suitability of those to their application, as well as identify areas for additional research. (C) 1997 Elsevier Science B.V.