Bayesian accelerated life testing under competing failure modes

This paper presents two models of Bayesian framework for the analysis of accelerated life testing (ALT) data with possible multiple failure modes. This approach does not yield an analytical solution, but allows small sample size and relies on the use of experts judgment in specifying the prior information in the inference. Gibbs sampling is used for posterior inference. Illustrative examples discussed in this paper show the applicability of the proposed models.