Google's DeepMind Develops Table Tennis Robot with Promising Performance

TapTechNews August 11th news, on August 7th, Google's DeepMind company announced that the R & D team has developed a table tennis robot that can reach the level of human amateur table tennis players in the competition.

According to the introduction, through 29 games between the robot and humans to evaluate the effect, where the robot won 45% (13/29). All the players are human players that the robot has never seen before, and their skill levels vary from beginners to championship levels. Although the robot lost all the games against the most advanced players, it won 100% of the games against beginners and 55% of the games against intermediate players, having the level of an amateur human.

TapTechNews noticed that as early as 2022, Google announced its ongoing research project of the table tennis robot (i-Sim2Real), which can catch the ball 340 times in one round when playing against humans. Google claimed at that time that with the speed at which AI grows, it will soon target professional players.

Google believed at that time that the advantage of table tennis is that it is rather strictly limited (compared to playing basketball or cricket), and it strikes a balance between complexity and simplicity. In this process, the machine learning model is taught what to do in a virtual environment or simulation, and then these knowledge are applied, with the goal of catching the ball rounds with humans for as long as possible without making mistakes.