Managing Hybrid Manufacturing/Remanufacturing Inventory Systems with Random Production Capacities

In this paper, we study hybrid manufacturing/remanufacturing inventory systems that produce a single product to satisfy random demands over a finite planning horizon. In each period, the firm receives random demand and random product returns. A serviceable product can be manufactured from raw materials or remanufactured from a returned product. Both operations face random capacities modeled as positively dependent random variables. The firm’s objective is to minimize the expected total discounted cost over the planning horizon. We partially characterize the firm’s optimal policy for the general model and completely characterize it for the models with deterministic manufacturing or remanufacturing capacity, by two increasing functions with slopes at most one. For the special case with unlimited manufacturing capacity, we further characterize the optimal policy and obtain additional insights. In particular, we connect this model with an auxiliary dual-sourcing inventory model and show that they have the same optimal policy under certain conditions. Finally, we conduct a numerical study to derive further insights into the effects of random capacities. Among others, we find that ignoring randomness of the manufacturing capacity often incurs significant cost to the firm while the cost of ignoring capacity correlation is negligible