The finals of China's first 3V3 (three versus three) robot soccer AI tournament (hereinafter referred to as “Robot Super League”) were held in Yizhuang, Beijing on the evening of the 28th, attracting 136 million online views.
The live commentary noted that while the robots were physically powerful, their “brains” were not yet clever enough, lacking coordinated passing, which reduced the entertainment value of the game. Analysts point out that humanoid robots still face challenges such as technology and cost, and the event economy will be a key driving force in technological development over the next decade.
According to China News Service, from performing the “yangge” dance at the Spring Festival Gala in the Year of the Snake, to the world’s first humanoid robot half marathon, enthusiasm for humanoid robots has been steadily rising. Unlike the robot half marathon, playing soccer requires robots to demonstrate achievements in areas such as motion control, visual perception, localization and navigation, and decision making.
After preliminary rounds, Beijing Information Science and Technology University’s Blaze Team, Tsinghua University Future Lab's Power Intelligent Team, Tsinghua’s Vulcan Team, and China Agricultural University’s Shanhai Team advanced to the finals. With the whistle sounding, the robots began their offensive and defensive displays, occasionally colliding during fights for the ball. Each goal was met with cheers and applause from the audience.
Humans Carry Stretchers to Rescue “Wounded Soldiers”
Comical scenes also played out frequently at the venue: robots fell either after physical “scuffles” or due to accidentally losing balance. Field staff would carry the toppled robots off the field on stretchers to minimize interruptions to the match. Some spectators joked that the stretcher crew were the busiest people at the soccer match. After three hours of play, Tsinghua University’s Vulcan Team won the championship.
“To enable robots to play soccer autonomously, we design algorithms to enable vision recognition to extract the positions of the ball and goal, use decision-making algorithms to select the right actions, and let the motion control system send instructions to the joints.” said Wang Yushi, captain of the Vulcan Team. The team drew on human team coordination experience to assign different robots to attract attention, defend, support, and shoot.
Self-Reliant AI Strategies in Operation
Competition Committee Executive Director Cheng Hao stated that all participating robots operated entirely on self-reliant AI strategies, with no need for human handlers’ intervention, and could get up after falling entirely by themselves. This “full mobilization” on the soccer field is a microcosm of the robust development of Beijing’s robotics industry.