Today for AI
HOT RADAR
agentHEAT 11.1°

AI CURATED HOT STORY · 10/30/2026

Tsinghua et al. Propose Collective Experience Learning for Robot Swarms

1 reports archived1 independent sourcesupdated 10/30/2026, 11:48:04
Synthesis & Latest Updates
1 Sources Cross-Validated

Tsinghua University, together with Infinigence and AgiBot, introduced a new paradigm called Collective Experience Learning (CEL) to address the scarcity of human demonstration data in scaling embodied AI. CEL enables robots to learn from peers' successful experiences rather than relying solely on human guidance. Validated on a cluster of 96 real robots, untrained agents achieved a 97.5% success rate across five complex tasks, surpassing average human operators. The study demonstrates a positive scaling trend where performance improves as the number of participating robots increases.

LATEST/Tsinghua University, together with Infinigence and AgiBot, introduced a new paradigm called Collective Experience Learning (CEL) to address the scarcity of human demonstration data in scaling embodied AI. CEL enables robots to learn from peers' successful experiences rather than relying solely on human guidance. Validated on a cluster of 96 real robots, untrained agents achieved a 97.5% success rate across five complex tasks, surpassing average human operators. The study demonstrates a positive scaling trend where performance improves as the number of participating robots increases.

Hourly heat curve

5 fully observed hours
Now
11.1
Peak
11.912:00
24h change
–

The line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.

Coverage timeline

Total 1 reports · Latest first
  1. 量子位官网T2·85 pts

    Tsinghua et al. Propose Collective Experience Learning for Robot Swarms

    • Introduced Collective Experience Learning (CEL), enabling robots to self-evolve by sharing and reusing peer successes, reducing reliance on human demonstrations.
    • Validated on a cluster of 96 real robots; untrained agents achieved a 97.5% overall success rate in five tasks (e.g., flower arranging, assembly), surpassing human averages.
    • Confirmed the 'more robots, stronger system' scaling effect: overall performance significantly improved as the number of learning robots increased from 16 to 96.