Belgian spacetech startup EDGX just scored a €2.3 million seed round to accelerate the rollout of its Sterna edge AI computer for satellites—a system designed to process data in orbit instead of beaming terabytes back to Earth for analysis.
The round, co-led by the imec.istart future fund and the Flanders Future Tech Fund (managed by PMV), comes alongside a €1.1 million multi-unit sale to a satellite operator. EDGX has already locked in its first space test: an in-orbit demo on a SpaceX Falcon 9 mission slated for February 2026.
From “Store and Forward” to “Process and Act”
Today’s satellites often operate on a “store and forward” model: capture raw data, store it, and send it down for processing later. But with Earth observation payloads generating petabytes of imagery and sensor readings, that approach is creaking under the weight of modern space missions.
EDGX’s Sterna Computer takes a different approach—run the AI models in orbit, on the spot. Powered by NVIDIA’s high-performance data processing hardware and EDGX’s own SpaceFeather software stack, the system can execute complex algorithms without waiting for a downlink. That means faster insights, lower bandwidth costs, and potentially mission-critical decisions made in near real time.
The SpaceFeather stack includes a space-hardened Linux OS, autonomous health monitoring, radiation fault detection and recovery, and the ability to deploy new capabilities post-launch—essentially making satellites upgradable like smartphones, but with better cosmic ray shielding.
Betting on AI-Powered Space Infrastructure
Founder and CEO Nick Destrycker says the market isn’t waiting for lengthy test cycles: “Customers aren’t waiting for flight validation, they’re signing now… This funding enables us to scale to meet demand for real-time intelligence from space.”
For investors, the bet is clear. “The space industry is hitting a fundamental bottleneck,” says Kris Vandenberk of imec.istart future fund. “We’re generating massive amounts of data in orbit but still using outdated architectures. EDGX is solving this by putting AI-powered edge computing directly in space.”
The Competitive Orbit
Edge computing in space is an emerging battleground, with startups and defense contractors alike racing to reduce latency between data capture and action. EDGX’s NVIDIA-powered approach puts it in the same conversation as US-based firms like Orbit Fab and Hypergiant, though its emphasis on modular, upgradeable software could give it an edge in keeping satellites relevant over multi-year missions.
If the Falcon 9 demo goes as planned in 2026, EDGX could help make “real-time space intelligence” less of a buzzword and more of a standard operating procedure.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
