Reliability assessment of man-machine systems subject to mutually dependent machine degradation and human errors

Abstract

Many man-machine systems experience machine degradation and human errors, which may be mutually dependent and have detrimental effects on system reliability. On the one hand, machine degradation will increase fatigue inducing conditions and result in more human errors. On the other hand, human errors usually cause shock loads on a machine and accelerate its degradation. Therefore, machine degradation and human errors aggravate each other. To model the mutual dependence, we develop a Piecewise-deterministic Markov process modeling framework, which can incorporate machine degradation and human errors to evaluate the system reliability. In the framework, the machine degradation is described by a multi-state model with a Semi-Markov process, where the times of transitions due to the mutual dependence are time-varying random variables; a mathematical model is developed to evaluate the human error rate under the effect of fatigue-recovery, where human errors occur according to a nonhomogeneous Poisson process. A Monte Carlo simulation algorithm is implemented to compute the reliability. The turret of a lathe operated by a worker is presented to illustrate the effectiveness of the reliability model.