Competing failure mode modeling in a Bayesian reliability assessment tool

This paper describes the implementation of a Bayesian procedure for the assessment of time-to-failure distributions. The procedure is part of the current analysis capabilities of a reliability assessment tool, which is designed to help engineers the generation of reliability assessments for products still under design or already in operation. These assessments are generated based on a combination of operational and test data, including data originating from similar but not identical components and systems, as well as engineering judgments. The Bayesian reliability assessment procedure is based on a competing failure mode model, which allows component and system-level reliability assessments to be performed in a single analysis, based on component- and system-level reliability data. This paper provides an overview of these modeling and analysis capabilities, and illustrates their use in a numerical example.