Lockheed Martin, USC Launch La Jument AI Satellite Payload with Ability to Enhance Images in Orbit
A nanosatellite built by Lockheed Martin and the University of Southern California (USC) was successfully launched into orbit on Jan. 13, 2022, where it will begin its mission, including taking photos of Earth and refining the images from space.
Part of USC’s Dodona satellite, the La Jument 3U payload launched from Cape Canaveral, Florida, aboard a Falcon 9 rocket as part of SpaceX’s SmallSat Rideshare Program.
This cubesat mission is a joint USC-Lockheed Martin satellite project, which space qualifies artificial intelligence (AI) and machine learning (ML) technologies through a series of in-orbit demonstrations using Lockheed Martin’s SmartSat™ software-defined satellite architecture.
The spacecraft’s La Jument payload uses SmartSat™ and represents the first instance that the full SmartSat™ stack has flown in space. SmartSat™ allows satellite operators to quickly change missions while in orbit with the simplicity of starting, stopping or uploading new applications.
The system is powered by the NVIDIA® Jetson™ platform built on the CUDA-X™ capable software stack and supported by the NVIDIA JetPack™ software development kit (SDK), delivering powerful AI at the edge of computing capabilities to unlock advanced image and digital signal processing.
La Jument will enable AI/ML algorithms in orbit because of advanced multi-core processing and on-board graphics processing units (GPU).
One app being tested in orbit will be SuperRes™, an algorithm developed by Lockheed Martin that can automatically enhance the quality of an image, similar to some smartphone camera apps.
SuperRes™upscales and improves the details within a photo taken by the satellite. A low-resolution picture is taken and the same image is upscaled to a higher resolution. The algorithm is able to intelligently fill in details in the high-resolution photo where details are essentially unknown. This process enables the exploitation and detection of imagery produced by lower-cost, lower-quality image sensors.
“SuperRes™ is pretty impressive,” said Lockheed Martin software engineer Shawn Nissley. “One of the interesting things about this project is we’re actually establishing a processing in space, so rather than move as much data onto the ground and then do all the processing on the ground, we’re actually able to have the spacecraft do some of that processing itself. The SuperRes™ algorithm is running on the spacecraft. It’s happening on orbit, so I think that’s something that’s pretty exciting.”
Lockheed Martin employees who worked on this project say this launch demonstrates the company’s capability to successfully rapidly fly a low-cost cubesat or smallsat. Project engineer Erik Nash Carlson believes it represents the future of technology demonstrations in space.
“With components and computers getting smaller each year, these low-cost smallsats are an easy way to get new technology in space, especially because these are so small. They can catch a ride on almost any rocket going to space,” Carlson said.
Carlson emphasized that the collaboration with USC was a huge part of the project’s success.
“Allowing the students to take ownership over the integration of this satellite has been great and there is truly something priceless about being in Cape Canaveral for a launch and seeing Lockheed Martin engineers working along USC students – who are the next generation of engineers – on a satellite that’s going to space,” said Carlson. “Their enthusiasm and approach to solving problems has made it great to work alongside USC.”
The collaboration between Lockheed Martin and USC won’t stop with the launch, though. Over the coming weeks, the teams will work together to ensure the satellite achieves its 180-day mission.
“We added a UHF antenna on top of our building in Valley Forge, so USC will operate their UHF antenna from the West Coast, we’ll operate our UHF antenna from the East Coast and then we’ll kind of collaborate back and forth,” said Nissley. “We’ll see the vehicle early, they’ll see the vehicle later, so we’ll coordinate uplinking commands and downlinking files between the two ground stations. So, that’s another area of collaboration that we have to work out to establish the interactions between our two organizations.”
Lockheed Martin’s Compass™, an advanced mission planning app, will be used to enhance imagery collection during La Jument’s mission. The Compass™ team developed a new Target Selection Visualization tool which makes it much easier to identify a targeted region to photograph. Using a map overlay, users can pick a point on the map within the satellite’s coverage area and the tool translates that into messages relayed back to the USC team commanding the satellite.
Looking forward, Lockheed Martin is planning to launch other La Jument satellites into orbit. They’ll be 6U CubeSats that incorporate research, including new SmartSat™ apps, sensors and bus technologies.