The Omron Robotics device is watching the opponent and predicts when it will have to react, for example, to a strongly hit ball. The robot has 3D cameras, sensors, and arm.

Artificial Intelligence trains table tennis with machine learning. This means that with every ball being played, the robot develops. Forpheus is able to assess the opponent’s skills so that he can adjust to his level.

The robot has two “eyes” and the arm is controlled by a 5-axis motor. Cameras analyze the movement of the opponent and the speed of the ball. Comparison with the human eye is not over – artificial intelligence with two cameras sees in 3D and can recognize objects. An additional sensor analyzes the movement of the ball.

This is the fourth generation of robots playing table tennis. The first model of Omron Robotics was shown in 2014. Since then, Forpheus has acquired a number of new features, the most important of which is the ability to serve with an extra arm. Together with the robot is provided a screen, mounted in place of the grid. Forpheus displays messages such as difficulty level. It can also … praise the opponent.

In the future, robots are supposed to play with each other. Those trained will be able to teach machines that are less advanced.