A stochastic model of a repairable item inventory system

the long version

An inventory system for repairable items is studied. A stochastic model which cooridnates procurement and repair decisions is developed. Special attention is paid to modeling the repairable item system so that the results derived in this report are applicable for Navy inventory management. Long-run distributions for both the ready-for-issue and the non-ready-for-issue stock and many useful measures of performance are determined. Uses of the information to analyze the critical factors in a repairable item system and to determine optimal values of the parameters are pointed out. Numerical examples of the calculations of the measures of effectiveness are presented. Finally, modifications of the model to include the zero attrition case and consumable items are pointed out. (Author)