Real-time Onboard and Remote Vehicle Health Management

We propose a real-time diagnosis, prognosis, and mission planning solution for complex systems. The proposed effort leverages QSI's real-time model-based diagnosis engine, TEAMS-RT, and UConn's expertise in the areas of signal processing, estimation and prognosis. The resultant solution will collect sensor data, process it, and assess system readiness in real-time. In the event of degradation or failure, the source will be isolated promptly, and its impact on mission evaluated. If the failure is in a redundant subsystem, the reconfigured system can continue with its mission, possibly in a degraded mode. Otherwise, the remaining life or extent of the failure will be estimated, and mission priorities and available capabilities matched to modify the mission or to select an alternative mission. A unique feature of the proposed effort is that the failure detection, isolation and reconfiguration modules can operate remotely over a low bandwidth wireless link. This can be utilized by ground support to assess system health and to shadow the onboard reasoner. More significantly, this tele-diagnosis capability will enable real-time health monitoring and on-demand diagnosis of legacy systems that lack onboard reasoners. The technology developed here has many space and commercial applications, including patient health monitoring, and aircraft and automotive diagnosis. ~000056 ~