Risk assessment on the EA-6B aircraft utilizing Bayesian networks

Military weapon systems often remain in service longer than anticipated. Systems must continue to operate safety and effectively while maintaining mission readiness. Degraders to readiness, such as high failure items, excessive repair times, long logistics delays, and manpower shortfalls, must be anticipated in order to proactively reduce risk. We applied a Bayesian network to a field data set obtained from the U.S. military. Our approach yielded a predictive method with substantial benefits over reactive methods, and was able to predict failure of several important components, to include potential malfunction codes.