markdown formatted blog content
The Problem with GPS in Space
Here's something most people don't think about: when we talk about satellites "navigating," they're usually relying on GPS signals from... other satellites. It works great when you're close to Earth, but take a trip to the Moon and suddenly you're in a navigation dead zone. GPS signals don't reach that far, and they get pretty sketchy even just a few hundred kilometers above our heads.
This is a big deal as NASA (and plenty of other space agencies and companies) start thinking seriously about lunar orbits, Mars missions, and swarms of satellites working together around other planets. You can't exactly pull into a gas station and ask for directions when you're drifting through the void.
So NASA did what NASA does — they found an elegant solution hiding in plain sight.
Your Trash Is My Navigation System
The Starling mission — a small constellation of four tiny satellites launched in 2023 — just demonstrated something pretty remarkable. Their FALCON system (which stands for something a bit of a mouthful: Fast Autonomous Lost-in-space Catalog-based Optical Navigation) can determine exactly where the spacecraft is by looking at other objects floating nearby.
Think of it like this: if you're in a dark room and you can't see the walls, but you can see a bunch of glowing objects around you, you could figure out roughly where you are based on where those objects are relative to each other. That's essentially what FALCON does, except the "dark room" is space, and the "glowing objects" include active satellites, dead satellites, rocket bodies, and general orbital debris.
The system uses the spacecraft's star tracker cameras — which are normally just there to help the satellite figure out which way it's pointing — and runs them in a completely new way. Instead of just looking at stars, FALCON identifies any object it can see, checks it against a public catalog of known space objects (yes, there's an actual catalog of space junk out there), and uses those confirmed objects as reference points to calculate its own position.
200 Objects in 3 Days — And Counting
Now here's where it gets really interesting. During testing, FALCON identified and tracked objects observed by Starling's cameras. It compared what it saw with the catalog data and then used those observations to figure out not just its own location, but also to improve the orbital predictions for the objects it was tracking.
Over just three days, without any input from ground controllers, FALCON refined the orbits of more than 200 different objects. And get this — the resulting estimates were actually more accurate than the original catalog data. The satellite essentially taught the system something new about its surroundings.
That feels like something out of science fiction to me. We've got spacecraft now that can essentially do their own astronomy homework and come back with better answers than what we had before.
Why This Matters for the Future
I'm genuinely excited about this for a few reasons.
First, think about space traffic management. We have thousands of active satellites up there now, and that number is growing fast. Being able to track objects without relying entirely on ground-based radar networks would make collision avoidance much more robust. If satellites can see each other and share that information, everyone benefits.
Second, and maybe more excitingly, this is exactly the kind of technology we'll need for those ambitious lunar Gateway plans and future Mars missions. When you're trying to coordinate multiple spacecraft around the Moon, you can't be constantly calling home to figure out where everyone is. The lag alone would make real-time guidance impossible. But if your satellites can navigate themselves? That's a game-changer.
The Starling team is planning to push this even further later this year. The four spacecraft will share their tracking observations with each other, combining their data to refine positions collectively. Imagine a little constellation of satellites all working together to map their own neighborhood in real-time. Pretty cool, right?
From Stanford Lab to Outer Space
I love that this started as a university research project. EraDrive, the company now commercializing the technology, emerged from Stanford University through NASA's SmallSat partnership programs. That's exactly how space technology should evolve — take good ideas from researchers, give them a real environment to prove themselves, and if they work, help them reach the broader market.
We're going to need a lot more smart software like this as space gets more crowded. GPS isn't coming to the Moon anytime soon, so we might as well teach our satellites to figure things out for themselves.
And honestly? If these little spacecraft can navigate using random bits of space debris as their guideposts, maybe there's a life lesson in there somewhere about making the best of what's around you.