DYNAMIC STRATEGIES FOR PREVENTIVE MAINTENANCE SCHEDULING WITH THROUGHPUT TARGET VARIATION

摘要

This paper considers a period-review production and preventive maintenance (PM) scheduling problem for manufacturing systems subject to uncertain demand in a finite-horizon. We consider multiple levels of PM which provide the system with flexibility of choosing various types of PM to be adaptive to the throughput target. A model with PM flexibility is proposed to simultaneously determine optimal production quantity and PM level for a single-product manufacturing system. A real option analysis (ROA) is applied to quantify the benefit of PM flexibility so that it can be considered in a cost function. The optimal strategy for a joint maintenance and production system that maximizes the overall expected profit of the system is obtained by a stochastic dynamic programming approach. Results are supported by numerical examples and show the comparison between the proposed PM-flexible model and the conventional PM planning model with single and fixed PM type. Managerial insights are provided on making adaptive PM and production decisions by trading off revenue and costs of corrective actions in different manufacturing applications.