Abstract
Planning, designing, and implementing a prognostics system requires a systematic approach stemming from business needs to system deployment. Since prognostics make a fit for service assessment of specific assets, the systematic approach begins with a service needs assessment. Given a specific asset or collection of assets, the level of abstraction is defined along with performance metrics. Next, an assessment of data driven, model driven, or hybrid approaches is reviewed to arrive at a prognostic methodology. With the methodology in place, sensors and data acquisition systems form the data acquisition plan. With the data acquisition plan, deployment and experiments are conducted to test and evaluate the prognostic system. Finally, a cost benefit analysis is performed prior to deployment and during testing to determine the solution feasibility, both technically and financially.
