6 Robots Build Great Wall in 15 Hours
新智元 · wechat · 2026-07-17
At WAIC, Yuanli Lingji showcased a public challenge to build a Great Wall out of 80,000 blocks. Without human remote control or pre-written scripts, 6 robots worked continuously for 15 hours to autonomously handle the entire process from grasping and transporting to assembly.
The article highlights the embodied model behind this demo, DM0.5: upgraded from the previous generation in data scale, architecture, and open-world generalization. It now features 60-second context memory, embodied CoT reasoning, and a trajectory alignment layer, achieving an SOTA score of 43% overall success rate and 54.42 composite score on RoboChallengeTable30V2.
The author interprets this challenge as a validation for the "last 20%" of precision manufacturing. Only when robots can collaborate, self-correct, and execute stably in the real world for extended periods can previously human-dependent tasks like precision assembly, flexible material handling, and rework be gradually automated.
Related event: 6 Robots Autonomously Build 80,000-Block Great Wall in 15 Hours(2 posts)→
More from Embodied
- RealSense D585 Pro AI-Powered 3D Stereo Depth Camera Unboxed for Physical AI — chrismatthieu · 2026-10-26
- Physical world agent swarm demo shared on X — pduan · 2026-10-26
- Looking for a wearable mic with best battery life that feeds recordings to personal agents via MCP — altryne · 2026-10-26
- 'Lamp Swarm': a desk of lamp-shaped robots shows what physical AI work looks like — dee_hw · 2026-10-26
- Tencent's HiSTEVE 1.0: hierarchical VLA agent plays Minecraft with native keyboard and mouse, hits 85% SOTA — _akhaliq · 2026-10-26
- SJTU's VaM lets robots learn online from mistakes, 95%+ success over 4-hour autonomous runs — chris_j_paxton · 2026-10-26