ABSTRACT
ABSTRACT: The optimization of the condition control of engineering systems which are comprised of a large number of homogenous components is considered. Since it is not feasible to inspect all components in such systems, the decisions for maintenance activities must be based on inspection results of a limited number of components. In this connection the decision criteria in regard to how much information is sufficient and in regard to which maintenance activities to pursue, are of particular interest. In the present paper, a Bayesian approach is proposed for identifying a criterion for how much information is sufficient and optimal to support decision making in regard to maintenance activities, employing the theory of sequential decision making. Two examples are given to illustrate how the proposed approach may be implemented into practical applications.
