Prognostic Methods for Service Life Management of Digital Electronics

We will provide an automated, prognostic decision making tool supporting Autonomic Logistics that predicts when digital electronics will fail to meet service requirements. Incipient fault-to-failure progression characteristics will be identified at the component and subsystem level to develop verifiable prognostic models driven by existing parameters and measurands. We will distinguish between normal equipment aging and slow failure modes. Accurate remaining useful life predictions will be facilitated and enhanced by our modular, open architecture that encourages use of multiple prediction and detection models, yielding highly accurate, timely, extensible, and flexible real-time PHM solutions. False alarms will be reduced significantly using our patented mode partitioning which recognizes distinct operating states and does not alarm when changing between modes. In Phase I, we will develop self-calibrating fault detection models that update dynamically and continuously to individual assets and to changes in individual assets over time. Useful life models and feasibility demonstration prototypes will be developed in the Option Phase with feedback provided by a top-tier JSF contractor. Proven software, system knowledge, and data gained from prior successful PHM USAF SBIR work will enhance and expedite our performance on this NAVAIR project. Self-funded, cost-sharing participation from an industry-leading electronics provider will enhance project success.