Cooperating expert systems for the next generation of real-time monitoring applications

A distributed monitoring and diagnosis system has been developed and successfully applied to real-time monitoring of interplanetary spacecraft at NASA´s Jet Propulsion Laboratory. This system uses a combination of conventional processing and artificial intelligence. Knowledge-based diagnosis modules are embedded within an automated monitoring system that detects on-board spacecraft anomalies. The diagnostic modules are specialized to respond to anomalies in a single domain of expertise and to cooperate with one another when necessary to solve complex problems that extend beyond an individual domain. Details of the distributed architecture, real-time diagnosis, and system performance are described in the paper. A brief summary of lessons learned in transferring research prototypes into operational environments is also reported