Simulation as a Method for Determining Inventory Classifications for allocation

Abstract

Companies that utilize multiple facilities to satisfy customer demand are faced with the same basic question - where should inventory be held? This thesis presents a method for answering this question, specifically for a company that allocates multiple units across multiple facilities, where any facility can fulfill an order to any customer, though with differing shipping costs. The model presented is a simulation of the shipping costs of various allocation strategies across a range of allocated inventory quantities, where the strategies simulated include consolidating all inventory in a central facility, constraining inventory to regional hubs, and spreading inventory throughout the network. The simulated results are then compared to find the low cost allocation strategy at a given level of allocated inventory. With this comparison, product groupings with the same low cost allocation strategy are identified, and are defined as "Slow", "Medium-A", "Medium-B", and "Fast" products. These groups can then be used to manage the allocation process, where "Slow" inventory is held centrally, "Medium-A" inventory held regionally, and "Fast" inventory spread throughout the network. "Medium-B" items serve as a costmitigating flexible option, where they are spread throughout the network when possible but consolidated when necessary to avoid changing the allocation for "Fast" items. At a broad level, the model presented is applicable to any company that can fulfill demand to a single customer from multiple facilities.