Aircraft Prognostics: Identifying Imminent Failures in Aircraft and System Components

We propose to develop software products for embedded, real-time monitoring of machinery health and failure prediction. The work will be based on new damage estimation, tracking, and prognostic algorithms that have been successfully developed in Phase I.Existing algorithms have been applied to data from the Boeing Reconfigurable Control and Fault and Identification System (RCFIS), an F/A-18 stabilator hydraulic control system test rig. Our algorithms can successfully discriminate between baselineconfigurations and those containing subtle faults, using only data readily available on aircraft: no new sensors are required. Phase II will continue to focus on software development for diagnosis/prognosis of aircraft hydraulic system grayscale healthmonitoring and failure prediction. Also, embedded applications in the industrial equipment and sensor industries will be developed. Additional tasks for Phase II include: defining the product in consultation with end users; enhancing the robustness andsensitivity of the damage tracking software; and better defining the data and processing requirements, and understanding how these are related to the specific task and signal environment. In addition to software products, this work will result in amethodology for evaluating the feasibility of specific applications, and for designing integrated diagnostics for embedded applications.