On-line inventory cost minimization for make-to-stock manufacturing systems

In this paper we study Make-To-Stock manufacturing systems and seek buffer capacities (hedging points) that balance inventory against stockout costs. Using a Stochastic Fluid Model (SFM), we derive sample derivatives which, under very weak structural assumptions on the defining demand and service processes, are shown to be unbiased estimators of the sensitivities of a cost function with respect to these capacities. Furthermore, the sample derivatives are nonparametric, i.e., not only are they independent of the distributions of the underlying random processes that drive the system, but they also require no knowledge of any model parameters (processing or demand rates). As a result, they can be evaluated based on data from real systems.