A Software Environment for Knowledge Discovery in Spacecraft Telemetry Data

P. Gonsalves, Alex Ivanov (Charles River Analytics), and P. Zetocha (USAF Research Lab)

Proceedings of the AIAA Space Technology Conference, Albuquerque, NM (September, 1999)

Ground operations stipulate significant program costs and manpower requirements for both commercial and military satellite systems. Spacecraft manufacturers and government agencies are constantly seeking methods to ameliorate these aspects of spacecraft command and control operations in light of budgetary constraints. Recent efforts to address this problem have seen a rise in the application of proven artificial intelligence and machine learning techniques into the spacecraft mission operations community to enhance automation and, in turn, decrease required ground support. Here we detail a software environment for knowledge discovery in spacecraft telemetry data. The developed system draws on the embedded knowledge, inferencing power, and learning and adaptability of intelligent software agents. The system provides a mechanism to automatically generate expert systems rules from spacecraft telemetry, thus reducing or eliminating the need for extensive and costly knowledge elicitation exercises.

System architecture

Request a Copy