ABSTRACT
In this paper we assume that the system cannot be continuously monitored but that inspections are possible allowing to determine the total damage incurred. We assume that there will be inspections at the times tn = nδ, so δ denotes the time between two successive inspections and inspections are not condition-based. In what follows we will mainly consider the case where critical failures are only detected during inspections. The case where the occurrence of a critical failure is detected immediately can also be handled, we will make some comments about the necessary calculations in the Conclusions in Section 7.
