A Cognitive Bayesian Toolkit for Payload Integrated Health Management Systems

MSI has a rich history of developing innovations in hardware and software for integrated health managemen with advanced Bayesian modeling and simulation.  This powerful technology can be adapted and enhanced for assessing payload performance, component capabilities, and situational self awareness.  We propose a new approach to sensor payload health management modeling that is based Cognitive Bayesian iNformation eXploitation,  learning of dependence as it combines and fuses data from in-mission collection of signatures and sensor data on the sensor system combined with probabilistic fusion with data about related systems (e.g. electrical power) and operating conditions. We propose to build on our success in using CBNX tools for engine health monitoring to develop CBNX for spacecraft payload integrated health management systems. In Phase I we will use the CBNX toolkit to develop and demonstrate a sensor payload subsystem prediction model. This could be an optical train to include cryogenic refrigeration and FPAs, gimbal tracking system or other sensor system. The model will be constructed based on information provided by the Air Force.    BENEFIT: We see this new toolset being used in any military application that benefits from a prognostic health management system. We see event larger markets for commercial applications including civilian space payload management and design.  There are extensions to civilian avionic system prognostic health management systems as well as any high value process system.