FA864924P0559

Small Technology Transfer Research Program (STTR) Phase II

Revolutionizing space-based ISR through decentralized systems & in-orbit ML computing for near-real-time intelligence

???????Intelligence, Surveillance, and Reconnaissance (ISR) capabilities are essential for effective defense operations. However, traditional space-based ISR systems suffer from high latency, limited data processing capabilities, and a lack of scalability, which can render the system fragile and vulnerable to security threats. As the amount of data and the number of space-based ISR assets continue to grow in the coming decade, these challenges will become more complex and difficult to manage. To address these challenges, we propose a Space-Based Distributed ISR System that combines edge computing and decentralized computing architectures to deliver resilience, speed (near real-time), security, and automation to space- based ISR. Our system leverages advanced artificial intelligence and machine learning models to extract intelligence from large amounts of raw data in real-time, avoiding the need for data transfer to ground stations. Our system utilizes a decentralized approach, which automatically coordinates tasking and intelligence generation with other satellites and assets over a secure, robust, resilient, and fault-tolerant network. By tasking a network of assets, instead of just a single satellite or constellation, we are able to provide a more comprehensive and timely response to events of interest. This system is designed to be compatible with a variety of assets, including satellites, aircraft, high-altitude pseudo-satellites (HAPS), drones, balloons, and space stations, providing a scalable and cost-effective solution. Instead of relying on a limited number of expensive aerial and orbital assets, we envision thousands of small, programmable intelligence nodes forming a distributed network. ???????Our proposed system offers significant cost efficiencies through a reduction in the need for data transfer and storage, as well as the investment in ground stations. Additionally, our system is designed to be expandable and scalable, with the potential to add cis-lunar and lunar capabilities in the future. We have already demonstrated our domain expertise in building intelligent nodes optimized for speed, size, and the space environment. Our single node demonstration as part of a decentralized system in space in December 2022 showcased a 96% reduction in time to generate intel, a 94% reduction in application size, and a 94% reduction in downlink data. We are also conducting projects to test various aspects of the proposed system, including real-time anomaly detection for monitoring space asset operations and onboard preprocessing of raw satellite images to remove undesired data and standardize the format. Overall, our proposed Space-Based Distributed ISR System provides an agile, AI-driven solution to enhance the effectiveness of DAF operations, allowing for quicker and more informed decisions that can significantly impact mission success.

The goal of phase II is to continue the R&D efforts initiated in Phase I. Funding is based on the results achieved in Phase I and the scientific and technical merit and commercial potential of the project proposed in Phase II. STTRs are completed in conjunction with a research institution.

University of Maryland

AFX23D-TCSO1

FX23D-TCSO1-0255

X23.D

Jenna Roeche