Optimal lot-sizing in an imperfect production system with homogeneous reworkable jobs

We consider a lot-sizing problem in a single-stage imperfect production system where the job processing is failure-prone. The processing may generate two types of defects: 1) reworkable defects which must be sent back for rework; 2) non-reworkable defects which are discarded immediately. The production process will switch between new jobs and rework jobs. Both new-job processing time and rework time are random. We discuss the optimal lot-sizing control, under a class of operating policies, to maximize the average profit over an infinite time horizon. We establish the existence of an optimal lot size. Furthermore, we characterize the profit function. Through the discussion, we obtain a very efficient algorithm for determining an optimal lot size.