A Comparative Study of Bayesian-Based Reliability Prediction Methodologies

As new technology is introduced into automotive engineering, the level of uncertainty regarding system robustness increases. With it reliability assessment tools that account for such uncertainty is expected to gain increased attention. This can naturally lead to Bayesian-based tools. This paper examines three reliability assessment methodologies that operate in the Bayesian framework. Two of them are geared towards electronic parts and assemblies, with the remaining one being geared towards systems in general. In doing so, they were critiqued in terms of four dimensions: (1) basic architecture, (2) input factors, (3) handling of qualitative data, and (4) failure rate updating mechanisms.