Prognostics and Health Management for Intelligent Mission Management

Digital electronic systems are pervasive in today’s aircraft, vehicles, weapons systems and satellites. Interface & Control Systems and Ames Research Center are teamed to bring crosscutting Prognostic Health Management technologies to the AFRL which can be applied to virtually any discipline with discrete and analog sensors. These technologies will be platform-neutral and available on embedded systems. Through Automated Machine Learning techniques, model-based reasoners will be populated and used to detect incipient failures and predict useful life remaining. The proposed architecture will replace labor intensive, error-prone, hand-generated logic with Machine Learning tools in an extremely cost-effective manner. ICS and Ames will demonstrate a highly automated model-based reasoning and rule-based expert system capability on digital avionics for both aircraft and satellite systems. Sensors will be characterized with signatures which will be monitored in real-time for deviations indicating a possible fault or failure. The rate of deviation from the signature will aid the Prognostic engine in determining the rate of failure for the device. The AFRL satellite programs also require the same technologies in support of the Intelligent Mission Manager for future autonomous missions such as the TacSat series of satellites.