An empirical study of an inventory-distribution system

This paper discusses the simulation results of a system consisting of a factory warehouse, and five regional distribution centers, each handling the same four products. The independent variables are demands, leadtimes, cost-rate structures. The basic data for this study has been taken from an appliance manufacturing company. Simulation experiments were conducted for three levels of each of the independent variables. The performance of the systems was measured against five important management criteria: average investment in inventories, average annual inventory carrying cost, average shortage cost per year, average reordering cost per year, and total number of reorders within the system per year. Analysis of variance was used to investigate the respective influence of each of the independent variables on each individual performance measure. The results of our study are new and of special interest, because they highlight the differences between the performance trends resulting from a single-item model and those resulting from a complex system, in terms of variations of the independent variables. This study proves that the trend of any performance-measure provided by a single-item single-location model should never be used to project that very trend of the aggregate performance of a complex inventory-distribution system.