On Saturday, SpaceX is launching a key piece of NASA technology that could one day allow satellites to repair other spacecraft already in orbit. It’s called the Raven, and it’s a module that’s going to test out a novel type of navigation system for space.
Because the module has three optical instruments, the engineers behind it have started calling it the “three-eyed Raven” — a nod to the psychic character beyond the Wall in the television show Game of Thrones. The reference is extra special given how the Raven is getting to space — on the back of a cargo capsule that’s launching on SpaceX’s Falcon 9 rocket. “The SpaceX capsule that is going to carry us is the Dragon,” Benjamin Reed, the deputy project manager for NASA’s Satellite Servicing Projects Division, tells The Verge. “So we’re the three-eyed Raven flying on the back of a Dragon.”
The new system is crucial in NASA’s quest to build a spacecraft that can autonomously meet up with other satellites already orbiting Earth. Specifically, the agency wants to create a special type of servicing satellite that can refuel or fix aging space probes. Typically, all satellites in orbit have a finite lifespan of about a decade, since they eventually run out of the fuel needed to operate them and they degrade over time. But a servicing satellite could potentially extend the use of satellites in orbit by many more years. Such a vehicle would act a bit like a traveling repair service in space.
“The paradigm for more than 99 percent of satellites in orbit is you operate your satellite, and when you run out of fuel, you build another,” says Reed. “If you have a robotic servicer to repair it or put in a new instrument, can you get more value out of that investment.”
The problem is that creating such a servicer is an incredibly daunting task. For one, the satellite has to be able to autonomously rendezvous with vehicles that are zooming around Earth at around 17,000 miles per hour. And since the servicing satellite will also be in orbit, it too will be traveling at the same ludicrous speed.
Plus, autonomous vehicles are something we’re only just now starting to master here on Earth — and self-driving cars have less degrees of freedom, or directional movements, to worry about than a vehicle in space. “For cars, it’s a ‘two degrees of freedom’ control issue,” says Reed. “That means [the car’s] controlling one of two things: the steering wheel left and right, or the gas and the brake.” In space, satellites have to be in control of six degrees of freedom. Like an airplane, a spacecraft can rotate three different ways; the satellite can also move up and down, left and right, or back and forth.
Because of this range of motion, no satellite has ever flown autonomously before, says Reed. In fact, a satellite’s position can vary quite a bit in space. Ground operators can guide a satellite’s movements in orbit, but the vehicle’s position can fluctuate up to several kilometers on its path around Earth. An effective servicing satellite, on the other hand, will have to be able to move with centimeter-level precision in order to catch up with another satellite, as well as reach out and grab it. And its movements will have to be autonomous due to communication delays between the ground and the spacecraft. Even a minutes-long delay can make things difficult when trying to fly with such precision.
This is where the Raven comes in: it will test the navigation system needed to make all of this happen. The module is a white box containing three key sensors — one that sees visible light, one that measures infrared, and another that uses a type of laser pulsing technology called LIDAR. These tools will work together to identify a vehicle in space and measure how far away it is, how it’s moving, and how it may be rotating. A future servicing satellite can then use the Raven-perfected system to figure out how to maneuver toward a spacecraft in need of refueling.
The Raven is among 5,500 pounds of food, supplies, and science experiments that the Falcon 9 rocket will deliver to astronauts on board the International Space Station this weekend. Most of this precious cargo will be riding to space inside the Dragon cargo capsule, but the Raven is traveling inside the Dragon’s trunk. That’s the unpressurized cylindrical structure that contains the Dragon’s solar panels and provides support for the capsule during launch.
Once the cargo capsule reaches the ISS, the Raven will be removed from the trunk by the Canadian robotic arm and attached to a platform mounted on the outside of the station. The Raven will stay there two years, analyzing the movements of cargo and crewed vehicles that come and go at the ISS. The module will track these spacecraft to see how good NASA’s navigation algorithms are. If there’s an issue with tracking, NASA can recode the Raven to make its navigation system more precise.
Reed and his team are eager to test out the Raven in space, because this type of navigation technology cannot be accurately tested on Earth. For one thing, the shadows in space are much different than shadows on the ground, which could potentially confuse the navigation system. “With there being no dust and moisture in space, the shadows from the Sun are super crisp, more crisp than they are on the ground,” says Reed. “We’ve written our machine driven algorithms to find the natural features on satellites, but we have to make sure our algorithms don’t get confused by these crisp shadows moving against a space object.”
If Raven proves successful, the next step for NASA is to try it out on an actual servicing satellite. The agency is working on a mission known as Restore-L, which calls for building a satellite that can refuel a spacecraft already in orbit. The Restore-L vehicle is supposed to launch in 2020, and will use the navigation system perfected by Raven to catch up to a global imaging satellite called Landsat-7. Not only will Restore-L refuel the satellite, but it will also grab hold of it and move it into a better orbit.
NASA’s goal for the mission is to create a new type of servicing capability that the private space industry can eventually capitalize on. That way, satellite operators will have more options in the future about what they want to do with their hardware in space. And that means potentially saving a lot of money. Large satellites can cost millions of dollars to build, and launching them to space is also a multi-million dollar endeavor. If a servicing satellite can go to orbit and repair up to 10 aging satellites, then the cost of launching that servicer is basically divided 10 ways. That means it could be more cost effective to repair a satellite than launch an entirely new vehicle.
“You can really get tremendous return on assets,” says Reed.