A dynamic maintenance policy for production system with variable operational condition

Abstract

This paper proposes a dynamic opportunistic preventive maintenance (PM) policy for a multi-machine system with batch production of pull pattern. The situation that the operational condition (OC) between each batch is variable and the information of next batch cannot be confirmed until the completion of current batch is studied. One objective of this study is to present the evolution model of reliability and hazard rate when OC changes, and develop a novel imperfect PM model of machine operating under variable OC. Setting that PM can be implemented only at the shift point of each batch, minimal repair is adopted to restore the machine back to work without altering hazard rate when failure happens during PM intervals, another objective is to propose a dynamic opportunistic PM policy to optimize the PM machine group at each batch shift point. A numerical example is finally introduced to validate the necessity of taking into account OC and to verify the effectiveness of our model.