Video: UC Berkeley humanoid robot plays table tennis with human-like agility

HITTER News

Video: UC Berkeley humanoid robot plays table tennis with human-like agility
HumanoidHumanoid Table Tennis RobotHumanoids
  • 📰 IntEngineering
  • ⏱ Reading Time:
  • 140 sec. here
  • 12 min. at publisher
  • 📊 Quality Score:
  • News: 87%
  • Publisher: 63%

Discover HITTER, a UC Berkeley humanoid robot that plays table tennis using AI-powered planning to outsmart human players.

UC Berkeley recently released a video demonstrating its latest creation, the Humanoid Table TEnnis Robot , playing a game of table tennis with human beings.The robot showcased exemplary skills with the bat, whacking the ball precisely in a rally of over 100 shots with human-like agility.

With its left hand outstretched, the robot used it to maintain balance while accurately playing shots, emulating human beings.Dual-system designHITTER’s performance is attributed to a two-part system that simulates how humans play.At the core is a high-level planner, which functions as the robot’s brain. It uses external cameras to track the ball in real time and forecasts where it will land. Based on this data, the planner calculates the precise location, speed, and timing required for a successful return shot.Complementing it is a low-level calculator that acts as the body. This system translates the planner’s calculations into coordinated robot arms and legs movements.Importantly, it has been trained on human motion data, allowing HITTER to move in ways that appear natural and intuitive. The robot can shift laterally, pivot, and swing its paddle with fluidity, reacting to balls traveling up to 5 m/s in less than one second.Training throughout reinforcement learningTeaching a robot to play table tennis presents unique challenges. The game requires split-second decisions, precise coordination, and adaptation to unpredictable shots. Berkley researchers addressed this by combining model-based planning and reinforcement learning.The model-based planner handles trajectory prediction and decision-making, while reinforcement learning fine-tunes HITTER’s movements through trial and error.This combination allowed the robot to return shots effectively and mimic human-style swings and footwork, making its movements surprisingly lifelike.HITTER could handle unpredictabilityHITTER’s framework was tested on a general-purpose humanoid robot platform, believed to be the Unitree G1, in live match conditions. During trials, HITTER sustained rallies of 106 consecutive shots with human opponents, a feat that would challenge most amateur players.The system also demonstrated versatility by competing against another humanoid robot, maintaining consistent exchanges across multiple rallies.According to the research team, these results show that HITTER can handle the unpredictability and speed of a real match rather than just controlled laboratory scenarios.AI-powered planningUC Berkeley’s HITTER is not just a curiosity in sports robotics; it represents a significant step toward robots interacting with the world in more human-like ways.By combining hierarchical planning, reinforcement learning, and motion data drawn from human players, the project provides a blueprint for building robots that are efficient and intuitive in their movements.As robotics continues to advance, systems like HITTER point to a future where machines are strong and capable and agile and adaptable, able to perform tasks with the speed and finesse once thought to be uniquely human.

We have summarized this news so that you can read it quickly. If you are interested in the news, you can read the full text here. Read more:

IntEngineering /  🏆 287. in US

Humanoid Humanoid Table Tennis Robot Humanoids Robot Plays Table Tennis Robotics Robots Table Tennis UC Berkeley

 

United States Latest News, United States Headlines

Similar News:You can also read news stories similar to this one that we have collected from other news sources.

Cecconi's bid to end Guardians' no-hitter drought cut shortCecconi's bid to end Guardians' no-hitter drought cut shortGuardians pitcher Slade Cecconi's was six outs from Cleveland's no-hitter since 1981 before the Royals spoiled the bid in the eighth inning.
Read more »

Tyler Glasnow dominates Rockies with no-hitter through six inningsTyler Glasnow dominates Rockies with no-hitter through six inningsLos Angeles Dodgers pitcher Tyler Glasnow has not allowed a hit to the Colorado Rockies through six innings on Monday night
Read more »

Dodgers flirt with another no-hitter, but this time they hang on to beat RockiesDodgers flirt with another no-hitter, but this time they hang on to beat RockiesTyler Glasnow throws seven no-hit innings, but Tanner Scott gives up a double to end the combined no-hitter in the ninth inning of a 3-1 win over the Rockies.
Read more »

Dodgers' no-hitter bid falls short in 9th, rally to beat Rockies 3-1Dodgers' no-hitter bid falls short in 9th, rally to beat Rockies 3-1The Los Angeles Dodgers have narrowly missed a no-hitter for the second time in three games
Read more »

Orioles Rally for Walk-Off Win After Holliday Breaks Up No-HitterOrioles Rally for Walk-Off Win After Holliday Breaks Up No-HitterThe Baltimore Orioles mounted a late comeback to defeat the Los Angeles Dodgers 4-3 on Saturday night. Jackson Holliday's two-out home run in the ninth inning broke up a no-hit bid by Dodgers ace Yoshinobu Yamamoto and sparked the Orioles' victory.
Read more »

The Sports Report: Dodgers lose no-hitter, but win game this timeThe Sports Report: Dodgers lose no-hitter, but win game this timeFrom Kevin Baxter: Years ago, when Don Drysdale and Sandy Koufax were at the top of the Dodgers’ pitching rotation, Drysdale missed a game to attend to some personal business.
Read more »



Render Time: 2026-04-01 19:05:17