Probability models applied to reliability and availability engineering

Our daily lives are dependent on various technological systems that may be mission-critical, safety-critical, or business-critical. Reliability and availability are crucial attributes and, thus, key requirements that should be considered during the entire life cycle of a technological system. Quantitative evaluation is an essential and fundamental task for ensuring and maintaining high reliability and availability throughout the system's lifetime. Since reliability engineering was established as a scientific discipline in the 1950s, diverse evaluation methods have been developed. This chapter provides an overview of major quantitative reliability and availability evaluation methods (measurement-based and model-based methods) as well as different features of systems studied in the reliability and availability engineering literature. An in-depth discussion is then dedicated to the model-based methods with the analytic-numeric solution. Among the model-based methods, non-state-space (or combinatorial) methods, state-space-based methods, and multi-level methods that combine the strengths of the prior two methods are explained and illustrated using examples. The application of the model-based methods in traditional power supply systems, as well as in emerging technological systems such as unmanned aerial vehicles and Bitcoin, are showcased.

Duke Scholars

Author Kishor S. Trivedi Electrical and Computer Engineering