Intelligent command and control systems for satellite ground operations

The Georgia Tech portion of the Intelligent Control Center project includes severalcomplementary activities. Two major activities entail thesis level research; the otheractivities are either support activities or preliminary explorations (e.g., task analyses) tosupport the research. The first research activity is the development of principles for thedesign of active interfaces to support monitoring during real-time supports. It well knownthat as the operator's task becomes less active, i.e., more monitoring and less activecontrol, there is concern that the operator will be less involved and less able to rapidlyidentify anomalous or failure situations. The research project to design active monitoringinterfaces is an attempt to remediate this undesirable side-effect of increasinglyautomated control systems that still depend ultimately on operator supervision.The second research activity is the exploration of the use of case-based reasoning asa way to accumulate operator experience and make it available in computational form.This project investigates the use of case-based reasoning technology to accumulate aknowledge base of actual operations experience and, subsequently, to use that experience asaid or advice in an intelligent decision support system. Initially, such a system willmonitor real-time operations, forming a knowledge base that reflects the range of nominaloperations. As unplanned and/or anomalous events occur the case base grows, in fact itautomatically learns, broadening its knowledge base to include operations experienceaccrued in managing these unanticipated events. Such a system uses case-basedreasoning technology to build an extensive repository of operations experience, i.e., cases,that over time, can function as the knowledge base for an autonomous system. This projectrepresents one of the first applications of case-based reasoning to real-time decisionmaking and system control. It provides an alternative, and potentially richer, knowledgebase than other AI techniques such as rule-based systems. Given the extent of operationalexperience that comprises the foundation of FOT expertise, a case-based system that canlearn from skilled operators is a promising way to encapsulate and capitalize on humanexperience and subsequently make it available to both other operators and intelligentsystems.Support activities for the two major research thrusts include the following:SAMPEX Operator Function Model, SAMPEX Task Analysis of Anomaly Detection andCorrection Processes, TPOCC Simulation/Testbed enhancements, and enhancements toTK+, an object-oriented toolkit for interface design and development.