Impact of the Frequency of Technician Visits On Facility Failure Rate

A 3-state Markov-chain model has been developed to analyze the relationship between the failure rate and the technician visitation rate for remote electronic equipment facilities. The major assumption of the model is that a facility, prior to failure, will give an observable indication that it is about to fail. In order for the model to predict the failure rate for a change in technician visitation rate, three parameters, obtained from facility maintenance logs, are used to initialize the model. These are the number of technician preventive maintenance visits, the number of pending-failures corrected as logged by the technician, and the number of hard failures. An upper bound failure rate prediction, for a decrease in facility visits, results from the optimistic assumption that the number of logged pending failures, if not corrected by the technicians, would have resulted in failures. The application of the model to two navigation-facility types indicates that changes in the visitation rate would moderately change the failure rate. Hence, the utility of the upper bound estimate is justified. Since the model is based upon a number of simplifying assumptions, it requires validation from a field test where the actual visitation rate would be varied, and the resulting failure rate would be analyzed. At present, a statistical design for the validation has been developed and is being seriously considered for implementation.