North America

Blue Room - Session 7

10:30 to 11:30 a.m. Thursday, June 15, 2017

A Practical Procedure for Bayesian Analysis of Reliability Data

Practical use of Bayesian reliability analysis is often hindered by the difficulties related to quantification of the prior information. Opposite to the exponential distribution, the Weibull distribution requires a two-dimensional joint prior. We propose a simple procedure, wherein the prior data can be taken in terms of a simple probability paper plot. Based on this and current test data, we derive not only posterior estimates of the Weibull parameters, but also the posterior estimate of the reliability function along with the respective credibility intervals. Real-world case studies are used as an illustration. We also show that the proposed procedure can be applied to any parametric lifetime distribution, not necessarily limited to the location–scale family.

Key Words: Reliability Data Analysis, Weibull Distribution, Location-Scale Distributions, Bayesian Inference

Vasiliy V. Krivtsov and Michael Frankstein

Ford Motor Company

Dearborn, Michigan