news
April 23, 2024
San Diego, California. The Pentagon is using artificial intelligence and machine learning to speed software improvements for unmanned vehicles across the services from months to days, Pentagon officials said Tuesday at the XPONENTIAL 2024 trade show.
While working on the Lionfish Unmanned Underwater Vehicle (UUV) program, operators realized they needed to update their software. To do so, he had to take the software out of the UUV, mail it to the engineers working on it, and then return it six times. A few months later, Doug Beck, director of the Defense Innovation Unit, said:
“One of the things the team discovered while working was that [with the Lionfish UUV] “That meant significant updates were needed to the software, from training missions to improved capabilities to improved search and evasion,” Beck said. Obviously we don’t have six months. ”
So the team deployed AI and machine learning tools to speed up the software improvement process. This allowed him to shorten the process from 6 months to just 1-2 weeks. In recent exercises with the Australians, that period was reduced to six weeks. day to day.
“This is a dramatic change, and it’s based on artificial intelligence,” Beck said.
Now, the Department of Defense is looking to expand that capability across the services so other programs can benefit as well.
“We are working with the chief digital and artificial intelligence officers on standardization,” Beck said. “We are looking at how to bring this functionality to other domains as well.”