Comparing scheduled and condition-based maintenance policies for single-unit systems operated in Markovian environments

We consider a single-unit system subject to a Markovian operating environment such that the system's instantaneous degradation rate depends on the state of the environment. The system fails when its cumulative degradation crosses some threshold. Using a simulation-based approach, we approximate the time to first failure distribution for this system with a Weibull distribution. We then develop an age-replacement policy for the system based on the Weibull approximation. As an alternative to the age-replacement policy, we also consider a condition-based maintenance policy. We compare the system performance resulting from the age-replacement policy with the performance resulting from the condition-based approach.