机器人界的“世界杯”:换台机器，冠军就换人
机器人圈第一次搞了场“世界杯”:同一套 AI 大脑，在 124 个真实场地、4 种不同机器人身上，打了 5699 场盲测。结果很扎心：没有全能冠军——在 A 型机器人上最强的模型，换到 B 型上可能垫底，排名一致性只有 0.24(满分 1)。更残酷的是，所有系统在开放世界里都很难稳定完成任务，连人形机器人上五个任务一个都没完成，最好的也只是“做到一半”。这告诉你：现在吹的“通用机器人”，离真正通用还很远，而且“在测试里赢”和“换台机器还能赢”是两码事。
📄 原文摘要(英文)
General-purpose robot policies must work across robots, tasks, and environments, yet the variation that makes physical trials informative also makes them difficult to reproduce. We introduce RoboColosseum, a distributed evaluation framework that combines unified policy serving and standardized protocols across a network of locally operated robots in diverse environments. Our initial study evaluates five vision-language-action (VLA) policies and three frontier AI agents across four embodiments (SO-101, Franka, Bimanual-YAM, and G1 Humanoid). The Open Track compares models without task-specific adaptation in blinded, pairwise trials under matched local conditions. The Fine-tune Track measures how well VLA policies adapt to 20 tasks across all four embodiments when each policy receives the same 100 demonstrations per task. Across 5,699 pairwise Open Track comparisons at 124 sites, frontier agents rival specialized VLA policies, while all systems struggle to complete open-world tasks reliably. After fine-tuning, GR00T N1.7 leads on SO-101 and pi0.5 leads on Franka and Bimanual-YAM. Policy rankings show weak agreement across embodiment-specific task suites (Kendall W = 0.24). Open Track rankings poorly predict relative performance after fine-tuning, with substantial rank reversals on all three embodiments shared between tracks. Measuring partial progress reveals differences that task success alone misses. On the G1 Humanoid, no policy completes any of the five tasks, yet average progress ranges from 13.3% to 23.2% across policies. Open Track rankings stabilize as trials accumulate, and active matchmaking achieves comparable ranking agreement with substantially fewer physical trials than uniform sampling. These results show how distributed physical evaluation can reveal where robot policies generalize, how well they adapt to new tasks, and how far they progress on tasks they fail to complete.