Robots are already participating in Olympic sports like running and high jumps, but they’re also showing progress in playing games requiring fine motor skills.
A robot developed by UC Berkeley researchers is able to play a table tennis rally with more than 100 shots with a human opponent. Named HITTER, or HumanoId Table TEnnis Robot via Hierarchical Planning and Learning, the robot autonomously tracks the movement of the ball, changes position, and hits its shots. Its movements are reasonably life-like, and it is able to effectively return balls at the level of a human beginner.
“Our humanoid robot can now rally over 100 consecutive shots against a human in real table tennis — fully autonomous, sub-second reaction, human-like strikes,” said Zhi Su, who was part of the team that made the robot.
Playing table tennis is tricky for robots, because it is a fast game, and robots must react to the ball within milliseconds. The robot not only needs to understand the ball’s trajectory, but also position itself to return it, and then perfectly time its hand with the exact right speed and angle to execute a shot.
“To address this, we present a hierarchical framework for humanoid table tennis that integrates a model-based planner for ball trajectory prediction and racket target planning with a reinforcement learning–based whole-body controller. The planner determines striking position, velocity and timing, while the controller generates coordinated arm and leg motions that mimic human strikes and maintain stability and agility across consecutive rallies,” the HITTER’s creators said.
Several companies are currently developing humanoid robots to work around the home and in factories. And while these robots can do things like folding laundry and working on factory floors, playing table tennis seems to be a bigger test of dexterity and skill — these research projects might not currently have many applications in the real world, but advances like these can help push the entire field of humanoid robots forward. And while the HITTER robot will likely not be winning Olympic medals anytime soon, it represents a significant step toward creating humanoid robots capable of excelling in fast-paced, skill-intensive tasks.