Engineering students have developed an AI robot that can beat them at foosball, adding another string to the technology’s growing bow.
At the end of the 1990s, AI had caught up with and overtaken humankind in the game of chess, with IBM’s DeepBlue first falling short against legendary player Gary Kasparov in 1996, before improving and beating him in 1997.
16 years later (back in March), AI had mastered an infinitely more complicated game called Go, with Google’s AlphaGo stunning world No 1 Lee Sedol with an emphatic 4-1 victory in the ancient game.
What’s next? Foosball, of course.
Anticipation and goals
Students at Brigham Young University in the US have been working on a robotic, AI-controlled foosball player that relies on cameras and quick movements. It’s already getting the better of some humans.
A camera set square above the halfway line monitors where the ball is, and where all the other ‘players’ are, controlling the figurines and anticipating, kicking and scoring.
“It’s becoming a challenge for us to beat the artificial intelligence,” said Nathan Warner, one of six students who created the AI.
“You think, ‘Oh yeah, humans ultimately should always be better than the computer’, but we’re actually struggling to keep up.”
Nice way to learn
Much like AlphaGo and DeepMind, the AI needs a backlog of moves to establish best modes of attack, so the students programmed in tactics they think humans use in foosball.
“Students learned how you can control machines to achieve the tasks that humans can do,” electrical and computer engineering professor DJ Lee said. “But, in this case, it actually reacts a lot faster than humans.”
Extrapolating this out onto other projects is the next step.