What are the effects of the reliability model uncertainties in the maintenance decisions?

Most of the works proposed for the design of reliability test plans are devoted to the guaranty of the reliability performance of a product but scarce of them tackles maintenance issues. On the other hand, classical maintenance optimization criteria rarely take into account the variability of the failure parameters due to lack of data, especially when the data collection in the operating phase is expensive. The objective of this paper is to highlight through a numerical experiment the impact of the test plan design defined here by the number of the products to be tested and the test duration on the performance of a classical condition-based maintenance (CBM) policy