Quality driven maintenance policies for a deteriorating system subject to non-self-announcing failures

Abstract

. In this study, we develop condition-based maintenance policies for a multi-state system subject to continuous-time Markovian deterioration that can result in non-self-announcing degradation including failure. Periodic inspections reveal the true state of the system; “good”, “medium”, “poor”, or “failed”. Upon receiving this piece of information, one of “do-nothing”, “minor repair”, “major repair”, “fix repair”, or “replace” maintenance actions is taken. On one hand, the system brings monetary rewards commensurate with system state–different rewards are earned per unit time spent in different states. On the other hand, quality loss that results from degraded system state is converted into monetary units–inspection and maintenance costs. Therefore, our objective is to determine the optimum inspection period and the corresponding maintenance policy that maximizes the expected long-run profit rate. We provide numerical examples to conduct sensitivity of the optimum inspection period and policy to system parameters, and to present the practical utility of our results.