Enhanced Prognostic Model for Digital Electronics

The ability to predict failures in aircraft electronic boards, their digital component elements and devices have the potential to reduce the risks of unanticipated failures while significantly reduce support costs. In this proposal, Intelligent Automation, Inc. (IAI) and Computer Aided Life Cycle Engineering (CALCE) Electronic Products and Systems Center (EPSC) at the University of Maryland propose an enhanced life consumption monitoring methodology for digital electronic boards and their components. Our approach involves a novel process to conduct Life Consumption Monitoring (LCM), including failure modes and mechanisms analysis (FMMA), virtual reliability assessment, sensor data pre-processing/feature selection, fault detection/identification/isolation, stress and damage accumulation analysis, and remaining life estimation. Meanwhile, the prediction output will be associated with a confidence distribution and adjusted by Support Vector Machine (SVM) and Confidence Prediction Neural Network (CPNN). Key advantages include better prediction of Remaining Useful Life than conventional methods, better prediction of some key parameters (thermal cycles and vibration loads) into the future so that prognostics information can be improved, incorporation of a simplified model that can provide "what if" predictions, and a data driven approach to improve the confidence of the overall predictions.