NASA's FALCON System Navigates Satellites Without GPS Using Orbital Landmarks

Satellites orbiting Earth have always leaned on GPS to know where they are. But GPS signals fade and fail the farther you travel from home—around the Moon, in deep space, anywhere beyond the reliable reach of ground transmitters. NASA has now demonstrated a different approach: a system called FALCON that lets spacecraft figure out their own location by looking at what surrounds them.

The technology, developed jointly by NASA and EraDrive (a startup born from Stanford University research), completed its first major test aboard the Starling spacecraft. Over three days, FALCON did something unprecedented—it not only determined Starling's own orbit using only onboard cameras and a catalog of known objects, but it also refined the orbital estimates of more than 200 other satellites and pieces of debris, all without a single command from operators on the ground.

The system works by turning spacecraft and orbital debris into landmarks. Starling carries cameras designed to track stars and bright objects in space—standard equipment on most satellites, used normally to help determine which way the spacecraft is pointing. FALCON repurposed these cameras for a new job: identifying objects in view, cross-referencing them against a publicly maintained catalog of known space objects, and then using those identified objects as reference points to calculate where Starling itself was located. The same observations that pinpointed Starling's position also allowed the system to refine its understanding of where those other objects actually were—improving on the ground-based estimates that had been loaded into the spacecraft's memory.

This capability addresses a real problem facing the future of space exploration. As NASA plans satellite swarms around the Moon, distributed science missions that require precise coordination across multiple spacecraft, and eventual human exploration of lunar and Martian surfaces, the need for independent navigation grows urgent. Ground-based tracking networks have limits. They can be congested, they require constant communication, and they don't work well when a spacecraft is on the far side of the Moon or traveling through deep space. A satellite that can navigate itself, that can maintain awareness of its surroundings and refine its own position without waiting for instructions from Earth, becomes far more autonomous and capable.

The FALCON demonstration also revealed something else: Starling's onboard predictions of where other objects would be turned out to be more accurate than the predictions supplied by ground stations. This opens a door to reducing dependence on Earth-based tracking infrastructure while simultaneously improving the ability to avoid collisions—a growing concern as the number of active satellites and debris pieces in orbit continues to climb.

The work itself traces a path from academic research to operational technology. What began as a university partnership project evolved into EraDrive, which is now commercializing the Era-Core software and hardware that made FALCON possible. Starling, which launched in 2023, provided the real-world testing ground. Later this year, the mission will expand the experiment: its four spacecraft will begin sharing tracking information with one another and using those combined observations to refine their collective position—a step toward the coordinated satellite networks that future deep-space missions will depend on.

Roger Hunter, program manager for NASA's Small Spacecraft and Distributed Systems program, called FALCON "yet another success" for Starling, noting that the implications reach across space-traffic monitoring, collision avoidance, and alternative navigation. The number of firsts from this mission, he said, keeps growing. What started as a question—can a spacecraft navigate itself by looking at what's around it?—now has a clear answer. The next question is how quickly this capability can be woven into the satellite networks of the future.