AgentCommand: A Control Center for Autonomous Distributed Satellite Systems

Agent-based systems, such as the ObjectAgent and TeamAgent systems developed by Princeton Satellite Systems, will facilitate the development of distributed, autonomous, and intelligent space systems. There are many advantages to using systems composed of multiple spacecraft, including greater performance, lower cost, and better fault tolerance, reconfigurability and upgradability. However, a significant disadvantage is the necessity of controlling many spacecraft simultaneously. Onboard autonomy, such as the Cluster Manager agent being developed for ObjectAgent/TeamAgent, partially alleviates this problem by creating a ¿virtual¿ satellite. In the end, however, it is still necessary for human users to interface with the system and this proposal addresses this critical area. A set of user configurable software tools will be created that enables users to easily command and monitor agent-based, multiple satellite systems, as well as perform fault detection, identification and recovery on the system. The tools will be internet-based. Although AgentCommand will be originally developed for use with the ObjectAgent software architecture, it will eventually interface with traditional flight software architectures and ground systems so that non-OA-based satellites can be included in the controlled formation.