Manufacturing learning and forgetting: Steady state optimal batch size for constant demand case

Abstract

Assuming learning and forgetting in processing units, constant demand rate, and infinite horizon, we analyze costs and properties related to lot sizes in the steady state. Steady State characteristics are described by a convergence in worker experience level or skills. The average per period cost as a function of lot size is found to be non-convex in the steady state. Thus, a simple approach such as first-order condition is not guaranteed to give an optimal solution. We develop sufficient conditions for existence of a uniqueoptimal solution, which are found in some problems. Our study shows that EOQ-type policies that use fixed batch size and produce when inventory reaches zero are not necessarily optimal.