Reliability and condition-based maintenance for multi-stent systems with stochastic-dependent competing risk processes

In this paper, we develop reliability and conditionbased maintenance models for a multi-component system of stents implanted in human arteries that is subject to both delayed and instantaneous failures. Considering the physics of failure and probability of failure, we develop a system-level reliability model and maintenance strategy by merging fracture mechanics and stochastic process approaches. In our new system-level reliability model, due to the exposure of stents to the same shocks, the component failure times are probabilistically dependent. Two patient groups with different activity levels are considered, and their corresponding system-level reliability functions are derived. We further develop a new condition-basedmaintenance policy, in which the replacement is either a preventive or corrective action performed depending on the condition of the system. Numerical examples using data from the literature are presented to analyze the reliability of implanted multi-stent systems and investigate the effectiveness of the proposed condition-based maintenance policy.