Conceptual Spaces for Space Event Characterization

In response to the DoD STTR 2022.D Solicitation for the Air Force Phase II Program, topic SF22D-T001, Advanced Space, LLC (Advanced Space) and the University of Texas at Austin (UT) propose to mature and customize hard and soft data techniques to characterize and identify Resident Space Objects (RSOs) and/or space events into a prototype software tool. The proposed solution features key capabilities that are combined into a Conceptual Spaces model. The Conceptual Spaces model characterizes space events with a threat concept similarity score to inform operator decision-making and Space Domain Awareness (SDA). The threat concept similarity score can range from expected station-keeping maneuvers (non-threat) to accidental or intentional kinetic impacts and to even a satellite rendezvous threat. With limited SDA assets available, this tool augments sparse data collection. Spacecraft travel around the Earth with tremendous energy, such that even the smallest objects can have lethal effects. Subtle differences in behavior can mean the difference between benign operations, high-energy impact, or espionage. Space event characterization is the process of parsing limited measurements of foreign object actions, estimating their intent and capability for potential threats and opportunities to harm, and determining the vulnerability of friendly spacecraft. Dr. Nathan Ré will be the Principal Investigator (PI) for the proposed project. He has an extensive background in low-thrust trajectory optimization and the use of neural networks (NNs) for autonomous operations of spacecraft. Dr. Ré is a Principal Engineer at Advanced Space. He has been PI and led more than 9 government-sponsored projects, totaling more than $14M in value. This track record includes an award by IARPA for the SINTRA program to detect, track, and characterize small space debris in LEO and another award by IARPA for the MicroE4AI program to apply ML to spacecraft navigation anomaly detection and classification. Advanced Space also supports the Joint Task Force – Space Defense Commercial Operations (JCO) in modeling and simulation software needs for the optimization of sensor tasking for SDA operations. This work primarily includes the creation of space domain events (e.g., spacecraft maneuvers or proximity events) for training and SDA exercises as well as data curation for threat detection operations. Mr. Charles Cain is the Advanced Space project lead for JCO work and will support this project to help guide the development towards SDA operations. Advanced Space will be supported primarily by Dr. Brandon A. Jones, who will be the Research Investigator (RI) for the proposed project. Dr. Jones is an Assistant Professor in the Department of Aerospace Engineering and Engineering Mechanics at UT. He has over 10 years of research experience in Bayesian estimation, uncertainty quantification, advanced numerical methods for astrodynamics, and related fields.