Oliver Schultz, Lockheed Martin; Joseph Brink, Lockheed Martin; Mark Williams, Lockheed Martin; Christopher Curry, Lockheed Martin
Keywords: analytics, automation, data integration, machine learning
Abstract:
The quantity and quality of data available to space operation centers is critical for operator understanding of the congested and contested space domain. The growing number of commercial Space Situational Awareness (SSA) data providers, advanced sensors, and expansion of common data repositories has increased the number of sources and types of SSA data that legacy systems were not originally designed to process. Decision makers need to quickly gain insights into the pedigree of the Space Domain Awareness (SDA) data feeds used in space battle management (SBM) systems before they can respond to activities in the space domain with confidence. Furthermore, additional sources of information do not always ensure that data is available to users in a timely manner. It is often not feasible for humans to manually understand the pedigree of SDA data because the lack of traceability and the volume of information is too large to interpret.
The research and development efforts conducted to enable space operators to quantify the pedigree and traceability of different SDA data sources are discussed. Automated analytical methods to quickly compare different sources of data and objectively rank data based on metrics such as traceability, latency, and pedigree are identified to help build trust. Additionally, this approach helps users to identify what process improvements would have the greatest impact to their mission. Spacecraft pattern of life machine learning algorithms, widely discussed in literature, can be enhanced with different datasets depending on the assessed pedigree. Examples show how these concepts can automate time consuming tasks reducing operator workload. Finally, the value of SDA analytical metrics, augmented with machine learning algorithms, running within space operations centers is highlighted.
Date of Conference: September 17-20, 2024
Track: Space Domain Awareness