Robots might be the new table tennis player, as DeepMind develops robots for playing table tennis. According to a paper named ‘Achieving Human Level Competitive Robot Table Tennis,’Google’s DeepMind Robotics showcased a robot playing Table Tennis.
The researchers explained that the robot was not a pro in the game. They further added that the robot can be a ‘solid amateur’ when played against any human player.
Play Table Tennis with a Robot
Google’s DeepMind posted on X (known as Twitter earlier) that “Robotic table tennis has served as a benchmark for this type. The robot has to be good at low level skills, such as returning the ball, as well as high level skills, like strategizing and long-term planning to achieve a goal.”
According to the paper, during testing, the table tennis bot was able to win over all the beginner-level players it faced. While playing with the intermediate players, the robot won about 55% of matches. On the other hand, the robot might not be ready to compete with pros. During the experiment the robot lost every time it faced an advanced player. Moreover, the system won around 45% of the 29 games it played overall.
The paper further claimed that Google’s DeepMind robot is one of the first robots to play Table tennis with humans. It also explained that this is just a small step towards a long goal in Robotics. The researchers believe that a lot of work needs to be done before getting the robot to achieve a human-level player capability.
The future ahead
Sports is expected to have been an important test for robots. One of the best examples of this phenomenon could be the annual RoboCup soccer competition, which dates back to the mid-1990s. However, according to the researchers, there are many shortcomings of using robots in the sports space. The biggest drawback of the robot is that it cannot react to fast balls. DeepMind explains that the reason behind this is system latency. In addition to this other reasons could be lack of useful data and compulsory resets in-between shots.
The researchers have also come up with a list of other drawbacks. This includes:
- Issues with high and low balls,
- An issue with the backhand
- It also lacks the ability to read the spin on an incoming ball.
Furthermore, “To address the latency constraints that hinder the robot’s reaction time to fast balls, we propose investigating advanced control algorithms and hardware optimizations,” the researchers highlighted.
Follow FE Tech Bytes on Twitter, Instagram, LinkedIn, Facebook.
