NASA Intern George Downey with homemade guitars. Credit: NASA/George Downey
exploration, moon, SCaN intern project, SIP, internship, intern, Profile 

NASA Intern Employs Machine Learning to Plan for Lunar Missions

Intern Innovation Among Homemade Guitars

By ​Danny Baird

August 25, 2020

During his virtual internship with NASA, Space Communications and Navigation (SCaN) program office intern George Downey developed predictive models that help NASA anticipate the needs of future lunar missions. His application of machine learning to NASA’s Artemis program will empower network engineers to provide robust and reliable lunar communications services.

Downey worked from his basement in Annapolis, Maryland, surrounded by a wall of guitars, a few amps, and a drum kit. His love of music ultimately led him to engineering. He built a few guitars, got interested in amplifier electronics, and then decided he’d rather “go to college than drop out of high school to join a band.”

Downey enrolled at the University of Maryland, College Park, pursuing a degree in electrical engineering. He’s currently an undergraduate senior who’s especially interested in communications hardware and circuit design.

Downey’s first experience with the SCaN Internship Project (SIP) was a shadowing day in 2019. There, he became interested in applying his engineering talents to space and met his current mentor, George Bussey.

After the shadowing day, Downey interned at the U.S. Naval Academy, working on the communications systems for a CubeSat using amateur radio bands. This summer, Downey turned from hardware to software, diversifying his work experience with SIP.

Downey classified current and planned missions by their science objectives, orbital dynamics, and data volumes to better understand their communications needs. He developed models that showed the network assets needed to support spacecraft, optimizing the program with machine learning. Ultimately, his models could predict the needs of a lunar mission with 83 percent accuracy.

Goddard engineers are now using Downey’s models to analyze the Deep Space Network’s (DSN) capacity to support lunar missions from 2024 to 2030. This is the first time the center has undertaken this work for the DSN — and it’s thanks in large part to Downey’s work.

While working on his summer project, Downey learned about the revolutionary communications technologies in SCaN’s innovation pipeline.

He pointed to optical communications as a particularly inspiring new capability for missions. Optical systems use infrared lasers to deliver far more data per second than comparable radio systems. Advancements in the technology will allow scientists to use higher-resolution instruments, enabling greater discoveries.

“It’s cool to see everything that can be done and has yet to be done,” said Downey. “Space communications has such a potential to advance technology. I’m excited by the possibilities.”

Downey’s work ensures that NASA’s networks are ready to support the Artemis missions. His models will be among the intern-developed communications and navigation innovations that empower a sustained presence on the Moon as NASA prepares for crewed missions to Mars.