Skip to main content

Tsinghua Wins Robot Soccer World Cup with ByteDance Tech

Tsinghua University Triumphs at RoboCup with ByteDance-Powered Robotics

In a historic achievement for Chinese robotics, Tsinghua University's Huo Shen team secured the championship at the 2025 RoboCup Humanoid Adult League, powered by cutting-edge technology developed in collaboration with ByteDance Seed Team. The victory marks China's first-ever win in this elite category of the international robotics competition.

Image

The Road to Victory

The competition, held annually since 1997 to advance robotics technology, saw Huo Shen dominate with scores of:

  • 6:3, 16:0, and 9:0 in group stages
  • 12:0 in semi-finals
  • 5:2 in the championship match

The team's success stemmed from their innovative use of decentralized computing and real-time dynamic response systems.

Breakthrough Technology: HumanoidKick Algorithm

The game-changing HumanoidKick algorithm represents a leap forward in robotic motion control:

  • Uses visual depth reinforcement learning
  • Enables complete action sequences from ball detection to kicking
  • Demonstrates superior adaptability compared to traditional methods
  • Allows real-time tactical adjustments during play

"This technology lets our robots respond to field conditions as fluidly as human players," explained Professor Zhao Mingguo of Tsinghua University.

Significance for Chinese Robotics

The victory establishes several milestones:

  1. First Chinese team to win RoboCup's Humanoid Adult League
  2. Validates China's growing expertise in autonomous robotics
  3. Demonstrates successful academia-industry collaboration between Tsinghua and ByteDance Seed Team
  4. Opens new possibilities for practical applications of humanoid robotics

Future Directions

The ByteDance Seed Team plans to:

  • Continue research into intelligent robot systems
  • Expand talent recruitment
  • Explore broader applications of their motion control technology

The collaboration aims to bridge academic research with real-world engineering challenges in robotics.

Key Points:

  • 🏆 Tsinghua's Huo Shen wins first Chinese RoboCup Humanoid Adult League championship
  • 🤖 HumanoidKick algorithm enables human-like soccer maneuvers
  • ⚡ Technology allows real-time perception and action execution
  • 🌍 Marks China's emergence as leader in competitive robotics
  • 🔬 Collaboration demonstrates value of industry-academia partnerships

Enjoyed this article?

Subscribe to our newsletter for the latest AI news, product reviews, and project recommendations delivered to your inbox weekly.

Weekly digestFree foreverUnsubscribe anytime

Related Articles

News

Inside San Francisco's Secret Robot Fight Clubs

An underground scene is electrifying San Francisco's tech circles - humanoid robots battling in steel cages while VR pilots control them remotely. These high-octane clashes combine Chinese-made hardware with American showmanship, supercharged by AI that makes the robots unnervingly lifelike. While thrilling audiences today, this emerging sport raises serious questions about where we draw the line between entertainment and ethics in robotics.

March 16, 2026
roboticsunderground techAI ethics
News

Sweet Potato Robot Raises $120M to Power Next-Gen Robotics

Di Gua Robot, a rising star in embodied intelligence, just secured $120 million in Series B1 funding. Backed by heavyweights like Synstellation Capital and Didi, the company is racing to develop full-stack robotics solutions. With chips, algorithms and software already under its belt, this Horizon Robotics spin-off aims to become the 'Wintel' of robotics. Their technology already powers everything from robot dogs to logistics bots through partnerships with industry leaders.

March 16, 2026
roboticsAI hardwareventure capital
News

Zeekr Unveils AI-Powered Mobility Agents at March Launch Event

Chinese automaker Zeekr is set to debut its 'Super Intelligent Agent' system on March 18th, featuring over 30 specialized AI assistants developed with Alibaba's Qwen model. These digital engineers promise to revolutionize smart mobility through collaborative task management, from solution selection to requirement distribution. The event marks Zeekr's latest push into intelligent transportation solutions.

March 13, 2026
electric vehiclesartificial intelligencesmart mobility
ByteDance Snags Alibaba's AI Talent Amid Industry Shakeup
News

ByteDance Snags Alibaba's AI Talent Amid Industry Shakeup

Yu Bowen, a key architect behind Alibaba's Qwen AI models, has reportedly joined ByteDance's Seed team following organizational changes at Tongyi Lab. This move highlights intensifying competition for top AI talent as companies race to develop advanced multimodal systems. The transition comes as ByteDance strengthens its visual and multimodal capabilities under former Google DeepMind executive Wu Yonghui.

March 12, 2026
AI TalentByteDanceAlibaba
News

Tech Talent Shuffle: Qwen's Key Players Jump to ByteDance

China's AI talent wars heat up as ByteDance snags another top mind from Alibaba's Qwen team. Yu Bowen, who led post-training for Alibaba's flagship models, joins ByteDance's Seed team in a move that signals intensifying competition in visual AI and multimodal tech. This comes amid broader restructuring at Alibaba's Tongyi Lab, highlighting how major players are scrambling to secure the brightest minds in foundational model development.

March 12, 2026
AI Talent WarsByteDanceAlibaba
Tencent and Tsinghua's AI Music Breakthrough Challenges Industry Leaders
News

Tencent and Tsinghua's AI Music Breakthrough Challenges Industry Leaders

Tencent has teamed up with Tsinghua University to launch SongGeneration2, an AI music model that's turning heads with its impressive 8.55% phonetic error rate - outperforming many commercial rivals. This innovative system combines language models with diffusion technology to create surprisingly natural-sounding music, even on consumer hardware. With its open-source approach and fast generation capabilities, it could democratize music creation.

March 10, 2026
AI MusicSongGeneration2Tencent