Bayesian reliability evaluation for customized products with zero-failure data under small sample size

Reliability information of some customized products is characterized as small sample and zero-failure. Traditional life time based methods and current zero-failure data based methods which require large sample size are no longer valid. As Weibull distribution has a wide application in describing failure times of various products, this paper aims to solve the problem for Weibull products. A new method is proposed to evaluate the reliability of products following Weibull distribution under the situation of small sample size and zero-failure data. To cope with the difficulties of small sample size, the proposed method tries to fuse the reliability data of both the target products and similar ones to enhance the accuracy of evaluation results. An example is used to demonstrate the validity and the effectiveness of the proposed method.