Most current research on optimal repair and replacement policies for machines operating in parallel does not utilize all of the information that is available. The purpose of this research is to provide repair and replacement policies that incorporate the dynamic information an engineer receives as the system is operating. The second major direction of this research is to investigate the situation where the parameter of an underlying exponential failure time distribution is unknown. A prior distribution for the unknown parameter is provided. This distribution is continually updated as information on the performance of the machines is gathered. A decision rule based on this updated information will be developed. The novelty of this approach is that it incorporates Bayesian statistical inference together with the probabilistic reliability methods that have traditionally been used.
