NASA SPAR Lab shares artificial intelligence tool for spacecraft

SAN FRANCISCO – Artificial intelligence promises to enable spacecraft to become increasingly resilient and able to collect data without waiting for instructions from ground controllers.

“We’ve been limited in what we’ve done so far,” said Evana Gizzi, artificial intelligence research leader at NASA’s Goddard Space Flight Center. SpaceNews“And there are so many things we want to do.”

For example, distributed missions, where spacecraft work together with landers and rovers to achieve common goals, will require autonomous capabilities. AI also paves the way for extensible mission architectures that allow new spacecraft and sensors to join orbiting swarms.

“Mission concepts at NASA and in the aerospace industry in general are becoming increasingly complex, which means many of them cannot be accomplished without AI,” said Gizzi, who earned a doctorate in artificial intelligence from Tufts University.

Methane Measurement

Still, incorporating AI into NASA missions isn’t easy. Space mission planners tend to be risk-averse and understandably wary of untested algorithms.

To reduce the barriers to introducing AI for spacecraft, NASA Goddard Space Autonomy and Endurance (SPAR) laboratory has created the Onboard AI Research platform called OnAIR. An open source software pipeline and cognitive architecture tool, OnAIR is publicly available on the software developer platform GitHub.

A prototype version of OnAIR was tested in NASA’s NAMASTE mission, which used a fleet of autonomous drones to measure methane dispersion in permafrost areas in Alaska. (NAMASTE stands for Network for Assessment of Methane Activity in Space and Terrestrial Environments.)

“OnAIR has helped drones maximize data acquisition in areas of high scientific interest by providing a standard for data collection and processing built into the NAMASTE software architecture. This standard also includes autonomous measurements made by the Multifunction Nanosensor Platform instrument integrated into the drones,” Mahmooda Sultana, who holds a doctorate in chemical engineering from the Massachusetts Institute of Technology and works as an instrument scientist at the NASA Goddard Planetary Environments Laboratory, said in an email.

Testing on the ISS

OnAIR has also been tested via SpaceCube Edge-Node Intelligent Collaboration, or SCENIC. SpaceCube is a family of commercially available, reconfigurable processors built with radiation-tolerant components. In 2023, a SpaceCube was tethered to the outside of the International Space Station.

After completing its primary mission, which included demonstrating the performance of commercial field-programmable gate arrays in space, SPAR Laboratory overcame many challenges to achieve the OnAIR demonstration.

“We initially thought we would have a full year to prepare and upload OnAIR to SCENIC. But about two months later, we learned that SCENIC would be decommissioned sooner than we expected,” James Marshall, a software engineer in the Science Data Processing branch at NASA Goddard who holds a doctorate in computer science from George Washington University, said by email. As a result, Marshall and his colleagues condensed a year-long project into six months.

They also learned to work with SCENIC’s slow main processor. “The clock speed was 100 (megahertz) and the architecture was less common, so it was difficult to port the Python libraries we needed and performance was slow,” Marshall said.

Connecting OnAIR to SCENIC’s existing core flight system presented another hurdle that the researchers overcame with knowledge from previous projects.

“The entire team (all three of us) had written software for SCENIC, so we were able to test everything on the SCENIC FlatSat and easily integrate the code,” Marshall said.

Relating to