Watch a Robot Show How Much You Suck at Foosball

The day foosball changes forever may be right around the corner.

Students and faculty at Brigham Young University just created an AI-assisted robotic foosball table. Initially intended to keep up with human players, the machine has far surpassed expectations — now the creators themselves struggle to beat it.

The device is built with an overhead camera and motorized foosball paddles. The students programmed the software to play how they assume humans would. Then using the camera, the computer calculates where the ball can go next based off color values of the paddles and ball, ultimately kicking it with lightning quick reflexes.

But why infuse a robotic foosball table with artificial intelligence? “It's not that we need a computerized foosball table, but it is a small example of a much larger problem,” says Nathan Warner, a computer engineering student involved on the project.

The exercise shows students what robotics and AI can achieve — no matter how big or small. “Through this project, students learn [the] concept: How you can control a machine to achieve the tasks that humans can do,” says professor D.J. Lee who oversaw the project.