The surge in low Earth orbit satellites in the near future, driven by the proliferation of mega-constellations, will create new avenues for malicious entities to disguise weapons or spy satellites amongst their fleets.
This warning comes from DARPA, the Defense Advanced Research Projects Agency, which has commissioned space startup Slingshot Aerospace to develop a new tool to identify potentially harmful satellites hidden in mega-constellations. The system, named Agatha after one of the “precogs” in the movie Minority Report, promises to significantly enhance national space security.
DARPA officially selected Slingshot for the Predictive Reporting and Enhanced Constellation Objective Guide (PRECOG) program last March, and the project wrapped up in January. The endeavor earned the company approximately $1 million, according to the government contracting database HigherGov.
Slingshot’s team created 60 years’ worth of synthetic constellation data to train Agatha, enabling the system to detect subtle discrepancies in satellite behavior and deduce the satellite’s actual operational purposes.
“There are many small clues that, when combined, reveal a much larger picture,” said Slingshot’s director of data science, Dylan Kesler, in a recent interview. These clues could include minor changes in satellite mass affecting its station keeping, unusual communication patterns with Earth, or consistent orientation in a specific direction compared to other satellites in the constellation.
Though Agatha was initially trained on simulations, it was later tested on real-world constellations by identifying non-threatening outlier satellites in existing fleets. Currently operational on Slingshot’s space domain awareness platform, the program now integrates data from the company’s Global Sensor Network, its Seradata database, and other public and proprietary sources.
While other startups have generated attention for their efforts to develop rendezvous technology to gather intelligence on adversarial satellites, Slingshot’s Agatha addresses the crucial initial step of identifying these satellites. This task becomes even more urgent as nations like China plan to deploy multiple mega-constellations in the next decade.
“Historically, when there weren’t these 10,000 or 15,000 mega-constellations, it’s understandable that humans or small teams could analyze orbital trajectory data and other information to make informed decisions on which satellites to target,” said Slingshot’s VP of strategy and policy, Audrey Schaffer. “But with the exponential growth and activity in space, it will become impossible for humans to effectively process all this data without tools like Agatha.”
