Abstract
Classical reliability modeling methods such as reliability block diagrams and fault trees express system reliability in terms of the reliability of the constituent subsystems and the architecture of that system. In recent years, prognostics and health management (PHM) has emerged as a promising method to combine sensing and algorithms to estimate important measures of reliability such as the probability that a subsystem possesses sufficient remaining useful life to conduct a mission without failure. This is especially important for mission critical systems. However, methods from classical reliability do not explicitly consider PHM. To overcome this limitation, this paper develops a modeling approach to consider reliability outcomes as well as PHM decisions, which should exhibit strong correlation in order to correctly classify the true state of the subsystem or component as healthy or unhealthy. We draw upon more general reliability modeling methods to characterize the correlation between the state of a subsystem's reliability and PHM decision. We subsequently propose an approach to obtain analytical expressions to assess system availability and cost in terms of these pairs of subsystem reliabilities and PHM decisions. Models that combine concepts from reliability and PHM will complement existing reliability, availability, and cost models, enabling sensitivity analysis within trade studies that can identify how improvements to subsystem-specific PHM techniques will impact system and fleet-level measures.