Reliability research of dependent failure: review and perspectives on modeling and analysis

Traditional system reliability research frequently relies on the classic assumption that component failures are mutually independent. However, in practical engineering systems, component failure processes exhibit significant dependence due to factors such as environmental coupling, load sharing, common-cause shocks, functional dependencies, cascading effects, shared resources, maintenance interactions, and degradation propagation mechanisms. Consequently, accurately characterizing these dependent relationships has become critical to improving the precision of system reliability assessments. This paper comprehensively reviews the research methods for dependent failure in system reliability analysis. We systematically categorize the connotations of dependence and critically evaluate representative modeling approaches for major dependent-failure mechanisms. Furthermore, this review synthesizes evaluation methods for dependent systems and analyzes the current challenges such as high-dimensional modeling and data scarcity. In addition, we examine AI-driven approaches for dependent reliability analysis. The review provides a clear framework for the theoretical research and development frontier of dependent failure in system reliability.