A Bayesian Dynamic Programming approach tooptimal maintenance combined with burn‐in

Abstract

We consider a complex system consisting of N identical components that are expected tofunction properly during a given period (mission time). Due to the possibility of failurebefore completion of the mission time, one allows for maintenance as well as burn‐in, thelatter in order to cope with the problem of “infant mortality”. We present a two‐level (Bayesian)Dynamic Programming approach to determine the optimal timing of the maintenance interventionsand durations of burn‐in, including maintenance‐induced burn‐in. We also discussways to reduce the “curse of dimensionality”.