Optimal recovery strategies for manufacturing systems

This paper presents a framework to generate optimal error recovery strategies in automated single-server assembly processes. The proposed methodology uses a semi-Markovian model to capture the stochastic nature of the production process when dealing with error recovery. The manufacturing objective in this paper is throughput or profit maximization. A machining process in which only a fraction of errors can be recovered is illustrated. Using the proposed framework, we determine the optimal policy for system operations and obtain the corresponding net earning rate or part throughput.