Hitch Open Ping-Pong Embodied AI Challenge

Table tennis is the ultimate test of a humanoid robot's speed, adaptability and sustained real-time decision-making. HOPE examines dynamic balance, high-speed motion control, real-time autonomous decisions and scene generalization in a single arena. It marks embodied AI's key leap from merely moving to competing and winning, backed by a quantifiable, reproducible open validation system.

No Remote Control
No Scripts
No Intervention
PerceivePredictPlanStrike

AGIBOT A3 humanoid closes the full loop independently

2026 Season

From its global debut at WAIC to an official event at the World Humanoid Robot Games, HOPE went from exhibition to sanctioned competition in a single season.

2026.7 · Shanghai

WAIC World AI Conference · Global Debut

At WAIC 2026, Hitch Open presented the AI autonomous-decision robot table-tennis exhibition, the first public showing of physical AI perceiving, deciding and executing in real athletic competition.

2026.8.22–26 · Beijing

World Humanoid Robot Games · Official Event

HOPE ran as an official competition event of the 2nd World Humanoid Robot Games at the National Speed Skating Oval, known as the “Ice Ribbon.” Twelve university and research teams competed on identical AGIBOT A3 robots, with no teleoperation and no cloud brains.

Champion
Peking University–BAAI Joint Team
View the Beijing standings →

2026 Participating Universities & Institutes

12 leading universities and research institutes on one stage

UC Berkeley
Tsinghua University
Peking University
Fudan University
Shanghai Jiao Tong University
USTC
University of Hong Kong
Shenzhen University
Beijing Information Science & Technology University
BAAI
Shanghai Innovation Institute
TeleAI, China Telecom

HOPE Global Standings · Beijing

Inaugural HOPE AI Challenge · 2nd World Humanoid Robot Games · August 2026

Tier: HOPE 250
Rank Team Result Points

Partners

Event Partners

  • 北奥集团
    Beao Group
  • 第二届世界人形机器人运动会
    World Humanoid Robot Games

Official Sponsors

  • 智元机器人
    AGIBOT
  • OptiTrack
    OptiTrack