WASHINGTON - Neuraspace, a space traffic management (STM) company based in Coimbra, Portugal, has introduced Machine Learning Prediction Plots, giving satellite and satellite constellation operators a tool for earlier collision avoidance planning.
The latest addition to Neuraspace’s STM software, using artificial intelligence (AI) enables operators to decide several days earlier whether to proceed with the available data or wait for additional data before making preparations for a collision avoidance maneuver. It gives them the means to decide if the data is good enough to make a decision.
As a result, operators, in particular those operating constellations with dozens or hundreds of satellites, have more decision time and can extend their satellites lifespan by saving valuable fuel and avoiding unnecessary maneuvers.
Neuraspace’s Machine Learning Prediction Plots calculate the path and forecast possible positions of the objects involved in a conjunction at the time of closest approach (TCA). Customers of the Portuguese company can access this information either through Neuraspace’s API or its website application.
"Neuraspace is the first STM company introducing “Machine Learning Prediction Plots," Chiara Manfletti, director and chief operating officer of Neuraspace, said. "Until now, no space traffic management tool was capable of making such an important forecast."
"Satellite operators already receive a deluge of alerts, most of them false, and therefore perform many unnecessary but costly maneuvers. A 300-satellite constellation may receive about 580 alerts, requiring human intervention and satellite maneuvers, per year. With an emergency maneuver in LEO costing about €25,000, this adds up to a staggering cost of €14 million per year. Saving some of these immense costs will make a huge impact."
Only made available last year, Neuraspace’s advanced space debris monitoring and satellite collision avoidance system is already being tested by some of the biggest satellite operators in the world.
