Data Driven Prognostics

The Airborne Laser (ABL) weapon system is one key part of a Department of Defense (DoD) approach to defending against ballistic missiles. The ABL is interested in developing a prognostic system that is able to provide an accurate picture of faults,component degradation, and predictive indicators of failures that will allow the operators to take preventive maintenance actions to avoid costly or catastrophic damage on critical parts and to maintain availability and readiness rates for the system. Theobjective of this proposal is to demonstrate the feasibility of addressing this problem using related past experience and a modular system developed by PulseMetrics as the general framework, and building on the technical expertise of the PrincipalInvestigators from PulseMetrics and Carnegie Mellon University in the areas of dynamical systems analysis, statistical techniques, signal processing and intelligent systems, to develop technology that will enable better health monitoring and lifeprediction for the ABL weapon system's critical components. There is an acute need for the development of a data driven prognostic system that provides advanced warning of failure, fault, and other error events in the context of the ABL system. Ingeneral, the ability to predict machine/equipment events has significant commercial potential in industrial applications where such capability would allow companies to improve reliability and safety, reduce downtime, and lower the direct maintenance costof physical assets. There is also potential use of this technology for pro-active, pre-emptive process intervention and course correction for operational and regulatory applications that support the re-vamping of industries such as the worldwide capitaland energy markets.