A comparison of allocation policies in a two-echelon repairable-item inventory model

The objective of this paper is to evaluate three allocation policies in a two-echelon machine repair model with state-dependent failure and repair rates. The first echelon consists of a number of bases that are served by a central depot in the second echelon. The operations at each base are modeled as a machine repair problem. Due to the complexity of the general model, a simulation study was conducted whereby the regenerative method of data collection was used. Two static and one dynamic allocation policy were compared with respect to five criteria. Based on the distribution of the observations, stochastic dominance tests were conducted and it was observed that static policies did not differ. The dynamic policy was found to be superior in a majority of the criteria.