In a series of test matches, the robot won 45% of the 29 matches it played against humans of varying skill levels.
The achievement is a milestone in the field of robotics and artificial intelligence. Google DeepMind researchers described their achievement in a research paper titled “Achieving human-level competitiveness in robotic table tennis.” The system combines advanced AI software with an industrial robotic arm equipped with a ping pong paddle.
The bot’s performance is impressive. In particular, it managed to win all matches against beginners and 55% of matches against intermediate players. However, it lost all matches against advanced players, indicating that there is still a lot of room for improvement.
The robot shows great abilities in the game. It can quickly execute a variety of forehands and backhands. In addition, it has the ability to read and adapt to the slight tilt that the opponent gives to the ball. It can also respond to balls coming at different speeds, from different parts of the table and at different heights.
Using table tennis as a demonstration area for robotic technology is nothing new. Several years ago, Japanese company Omron introduced a table tennis robot that could play continuously with a human player, demonstrating the capabilities of its sensors and control systems.
However, as artificial intelligence advances, table tennis robotic systems have made great strides. According to its creators, Google DeepMind’s robot represents a step toward achieving human-level speed and performance in real-world tasks, a goal that has been a challenge for the robotics research community.
-
1
More Stories
In Greece Porsche 911 50th Anniversary – How much does it cost?
PS Plus: With a free Harry Potter game, the new season begins on the service
Sony set to unveil PS5 Pro before holiday season – Playstation