Application of Bayes statistics to reduce sample-size, considering a lifetime-ratio

Abstract

Due to short development schedules, the amount of time available for testing is often unequal to the required lifetime of the product. Either a reduction or increase of the test time affects the confidence level of the reliability test. In this paper, the so-called "lifetime-ratio" is introduced for the Kleyner et al. method and for the Bayes approach, which uses a uniform distribution as prior knowledge. With the lifetime-ratio, it is possible to consider a case where the test time is unequal to the projected lifetime. Apart from the simple case of the success run, the authors also took into account that failures may occur during the test. In addition, they showed that the lifetime-ratio is dependent on the number of items that must be tested for a required reliability and on the desired confidence level.