Artificial intelligence shows vast potential for tackling old space problems
Artificial intelligence and machine learning will revolutionize the space industry, allowing users to cull through massive sets of data to find clear answers to problems from intelligence and climate analyses to traffic management in increasingly crowded orbits, a panel of experts told an audience at the 39th space Symposium.
Chiara Manfletti, the CEO at space traffic management firm Neuraspace, said artificial intelligence is better suited than people at tackling some space problems.
“Our vision is that if we want to unleash the potential of space going forward … we want to take the human out of the loop,” she said.
For now, though, computers are trying to tackle “niche” problems, finding clearly defined answers within data, said Mark Eddings, space market senior vice president at LMI, which creates software for the national security space sector.
“Within LMI we have our customers primarily on the government side just need quick wins,” he said.
But the small-scale victories for artificial intelligence hint at vast future potential. “These little wins can be transformative,” he said.
That potential is drawing massive investments, said Joe Schurman, consulting giant PricewaterhouseCooper’s principal leader for U.S. space programs, aerospace, defense, and artificial intelligence.
“Our firm invested over $1 billion in AI over the past year,” he said.
AI uses computing power to recognize patterns, which make it applicable to industries from national security space to public accounting.
“The use cases are in the thousands,” Schurman said. “It’s pretty insane.”
In space, customers are seeing value in artificial intelligence for situational awareness.
Manfletti said is adept at scanning huge data sets on satellites and space junk and developing collision avoidance. That would allow larger constellations to safely share orbits.
But new technology must overcome old problems before widespread adoption, Eddings said.
“Your biggest obstacle is security, security, security,” he said.
The security issues include developing trusted software that fends off hacking threats and keeps massive sets of sensitive data secure.