Full Breakdown
Loft Orbital and NASA JPL Test Edge AI on Satellites for Real-Time Earth Observation
6/24/2026, 11:54:30 AM
AI-Enabled Tip-and-Cue Test Launched
Loft Orbital and NASA’s Jet Propulsion Laboratory have begun testing JPL’s AI software on a Loft spacecraft under the NASA-funded Federated Autonomous Measurement (FAME) project. The first flight launched in June, with onboard computers running AI models to automate tip-and-cue, directing one satellite to capture detailed observations from another.
Background: Tip-and-Cue to On-Orbit
Traditionally, tip-and-cue requires raw images to be downlinked, analyzed on Earth, and commands sent to a second satellite—a process that can take hours. Embedding AI at the edge lets a spacecraft recognize features of interest in real time, flag them, and send concise alerts without bulk transmission.
Technical Hurdles
The program faces two challenges: integrating sensors and processors for real-time image analysis, and securing open-source AI models small enough for satellite hardware. Recent advances in multimodal, high-performance models and a “very large corpus” of training data now make these constraints tractable.
Impact on Monitoring
If successful, the system could deliver near-instant detection of wildfires, marine-pollution events, and other hazards. Lasserre also cited security, military, and intelligence uses where rapid situational awareness is critical. Real-time autonomy could raise the commercial and governmental value of space-based Earth observation.
Official Statements
Loft officials describe the effort as a shift from bulk downlink to edge-derived intelligence, noting that prepared state-of-the-art models can run on the satellite. The envisioned “patrol mode” would keep a sensor active, use AI to flag targets, and employ inter-satellite links for follow-up, creating a “tipping point” where autonomy increases data utility for customers and agencies.
Conflicts & Gaps
The available sources provide a single perspective; no independent verification of test outcomes or performance metrics such as detection latency or orbital model accuracy has been reported.
Verbatim Quotes
- “You can use different assets with processing at the edge to capture, sense, understand and send insights about what’s going on on Earth without having to downlink big amounts of data,” — Paul Lasserre, General Manager for AI, Loft Orbital
- “It can recognize everything without being told what to look for,” — Paul Lasserre, General Manager for AI, Loft Orbital
- “We can run state-of-the-art models if we have prepared them well.” — Paul Lasserre, General Manager for AI, Loft Orbital
- “You can have this ‘patrol mode’ that was not really a thing before” because of processing and connectivity bottlenecks, he said.” — Paul Lasserre, General Manager for AI, Loft Orbital
What’s Next
Loft plans AI-enabled flights in 2027 and 2028, followed by a 10-satellite Altair constellation with multiple sensors, edge computing, and inter-satellite links. Ongoing collaboration with JPL will focus on model compression, sensor expansion, and operational demonstrations.
