Abstract
Although the literature abounds with inventory models developed for almost every conceivable set of circumstances, when it comes time to develop an inventory system for a particular application, none of the literature models seem to fit exactly. This requires the analyst to do one or both of the following:a) Modify an existing model to fit the given situation.b) Assume that even though not all of the assumptions in the existing model are satisfied, it will do an adequate job of controlling the inventory.Frequently, simulation is used to evaluate the consequences of following approach b (see, for example references 1, 3, 5, and 8).The purpose of this paper is to demonstrate the use of simulation in the design of a forecasting and inventory control system. The use of alternate approximate modelling strategies will be simulated using an analytic simulation model of the real system on a relatively large sample of parts. The design questions under study will include the forecasting model, the frequency of updating the forecasts, the order quantity model, and the reorder point model. In each case, a very simple and a more complex model will be used. The major thrust is to point out the use of simulation as an aid in evaluating design alternatives.
Publisher
Association for Computing Machinery (ACM)