Data Mining Applications for Space Mission Operations System Health Monitoring

Recent developments in data mining techniques for anomaly detection make it possible to use the wealth of available archived spacecraft system data to produce advanced system health monitoring applications. These “data driven” applications are capable of characterizing and monitoring interactions between multiple parameters and can complement existing practice to provide valuable decision support for mission controllers. Data driven software tools, including Orca and the Inductive Monitoring System (IMS), have been successfully applied to mission operations for both the Space Shuttle and the International Space Station. Orca uses a nearest-neighbor approach to search for unusual data points in multivariate data sets by calculating the distance of each data point from neighboring points. The IMS tool uses a data mining technique called clustering to analyze archived spacecraft data and characterize nominal interactions between selected parameters. This characterization, or model, is compared with real time or archived system data to detect off nominal behavior. Augmenting traditional mission control software with advanced monitoring tools, such as Orca and IMS, can provide controllers with greater insight into the health and performance of the space systems under their watch. We will describe how such techniques have been applied to NASA mission control operations and discuss plans for future mission control system health monitoring software.