Confidence interval for optimal preventive maintenance interval and its applications in maintenance planning

Replacement problems of deteriorating systems have been extensively studied. Typically, the time between failures is characterized by lifetime distribution in which the parameters are estimated from historical data. On the other hand, in most cases, the work focuses on determining the optimal replacement schedule by assuming that model parameters are constant. Here, the issues arising from the use of estimated parameters are studied and the results are applied to opportunistic replacement. Also, a graphical approach is proposed to obtain the confidence limits for the optimal replacement time, considering the key parameters of the two popular replacement models, namely, the age replacement model and the block replacement model. The applications of the proposed confidence interval are presented, namely, determination of the window of opportunity for minimum cost, maintenance scheduling given erratic customer demand, and opportunistic maintenance for multi-component systems.