A Bayesian Scheme for Sequentially Testing a Multi-Component System

Although the literature on reliability theory is filled with methods for optimally designing multi-component systems subject to various constraints, very little is available in the literature on methods for optimally testing such systems. Both capabilities are needed, because after a system is assembled, periodic testing is required to insure against degradation of component reliabilities. Some components in the system should be tested more than others, because of their reliability, cost of testing, or placement in the system. This paper provides Bayesian decision rules for testing components sequentially to minimize the sum of the cost of testing and the cost of misestimation.