SpaceWERX Physics-Constrained AI Cuts Space Object ID Time by 99%

The U.S. Space Force’s innovation arm, SpaceWERX, has awarded BosonQ Psi Federal LLC (BQP) its first federal contract to develop a new approach to identifying the thousands of objects detected daily by the U.S. Space Surveillance Network. Between 18,000 and 25,000 observations are collected each day, and a significant portion remain unidentified, posing challenges to space situational awareness. BQP will build a Physics-Constrained Quantum-Assisted Machine Learning (PC-QAML) software application designed to accelerate and improve the accuracy of object identification in orbit, reducing AI model size by 99% while maintaining over 99% classification accuracy. “Our goal is to make advanced AI practical where it matters most: on satellites and forward-deployed systems operating with limited computing power and intermittent communications,” said Rut Lineswala, Founder and CTO of BQP.

BQP Secures SBIR Contract for Quantum-Assisted Space Domain Awareness

This funding validates BQP’s technology and establishes the New York-based company as a player in the U.S. federal market, fostering collaboration with stakeholders focused on understanding the space environment. Unlike conventional AI approaches reliant on cloud computing or powerful GPUs, BQP’s solution is engineered for deployment on resource-constrained edge devices, like space-qualified processors. This efficiency translates to a tenfold reduction in inference latency and approximately 90% lower power consumption compared to traditional machine learning methods, as demonstrated on an NVIDIA Jetson Nano at the BMC3I TAP LAB. This capability is particularly crucial for Space Domain Awareness, where rapid distinction between routine activity and potential threats, such as satellite maneuvers or electronic interference, is paramount.

The technology builds upon prior work with the Space Domain Awareness Tap Lab and supports objectives for both Space Operations Command (SpOC) Mission Delta 2 and Space Systems Command (SSC); it has already shown promising results in orbital separation detection during the 2025 SDA Mini-Accelerator. Beyond defense applications, BQP anticipates commercial uses for PC-QAML in areas requiring high-performance AI on compact, low-power hardware, including autonomous systems and industrial monitoring.