Abstract
This study focuses on a capacity rationing problem for make-to-order manufacturing, in which the allocation of finite capacity to various random inquiry orders from Poisson processes has to be determined. In make-to-order manufacturing environments, future orders with varying ready dates, due dates, processing times, and profits arrive randomly. With the objective of maximizing profit, a manufacturer should make the acceptance-or-rejection decision immediately when an order request arrives. If the manager accepts an order, the remaining capacity is reduced and the profit of the order is earned. On the other hand, if the manager rejects the request, the capacity is reserved and could be probably utilized to produce higher profit orders arriving in the future. This study focuses on such a stochastic rationing decision problem. A dynamic decision method called simulated expected revenue decision procedure (SER) proposed by Hung and Lai (2010) is adopted in this study to deal with the capacity rationing problem. This study conducts the simulation experiments to compares the results of SER, first-come-first-served (FCFS), and a decision criterion based on loading estimation. According to the experiment results, SER outperforms FCFS by an average profit of 3.98%. Besides, under various problem conditions, the performance of SER is more robust than FCFS. Therefore, SER is an effective decision method for the MTO capacity rationing problem. Keywords: capacity rationing; Poisson process; decision support system.