论文arXiv 2611.251052 天前
MAWAM多智能体世界模型实现协同操作67.8%实机成功率
本文由 AI 根据公开内容整理,可能有误,请以原文为准:MAWAM: A Multi-Agent World-Action Model for Coordinated Manipulation(arXiv 2611.25105)
核心内容
针对多机器人协同操作中环境动态耦合难题,研究提出MAWAM多智能体世界-动作模型。该模型基于预训练单机骨干网络,通过跨智能体注意力机制联合去噪所有机器人的动作token与未来视图token,并在单次采样中同步生成可执行动作块与各机器人独立控制接口。在洗衣袋收集、浴巾折叠及雨伞交接三项真实机器人任务中,MAWAM取得67.8%的平均成功率,比相同演示数据训练的MA-VLA高出18.9个百分点;在包含最多四臂团队的八项仿真任务中,平均成功率达90.3%。
消融实验证实架构设计有效性:移除跨智能体注意力后,四臂接力任务成功率从95.0%骤降至39.0%;缺失团队阶段未来视频监督时,Place Food任务成功率由78.0%跌至47.0%。这表明在多机协同场景下,仅靠单机预测无法处理队友行为引起的环境变化,必须将团队联合生成的未来物理状态作为显式约束。MAWAM通过将世界模型扩展至多智能体维度,解决了传统VLA在多机交互中因忽略队友影响导致的规划失效问题,为复杂具身协同提供了新范式。
关键看点
- 实机性能突破:在洗衣袋收集等三项真实协同任务中获67.8%平均成功率,较同数据MA-VLA基线提升18.9个百分点
- 联合生成机制:基于单机预训练骨干,利用跨智能体注意力联合去噪动作与未来视图token,单次采样输出多机动作块
- 关键模块验证:移除跨智能体注意力致四臂接力成功率从95.0%降至39.0%,证明队友感知是协同核心
- 视频监督价值:团队级未来视频监督使Place Food任务成功率从47.0%提升至78.0%,强化物理一致性约束
工程与产业观察
MAWAM证明了在多机协同具身智能中,构建共享的世界模型比独立决策更具优势。其通过跨智能体注意力与团队级视频生成监督,有效解决了多体交互中的环境非平稳性问题。该工作提示产业界,在开发多机器人系统时,应优先探索具备联合时空预测能力的端到端模型,而非简单堆叠单机策略,这对仓储分拣、柔性装配等强耦合场景具有重要工程参考价值。
英文摘要 (Abstract)
Every move in a robot team changes the world in which its teammates must act. Coordinated manipulation therefore calls for predicting the team's actions together with the physical future they jointly create. We introduce MAWAM, a multi-agent world-action model that jointly generates each robot's future views and executable action chunks in a single sample. Built from a pretrained single-robot world-action backbone, MAWAM jointly denoises all robots' action and future-view tokens through cross-agent attention, while preserving a separate control interface for each robot. MAWAM achieves the highest success on all three real-robot tasks (laundry-bag collection, bath-towel folding, and umbrella handover), with 67.8% mean success, 18.9 percentage points above MA-VLA trained on the same demonstrations and cameras. In simulation, MAWAM reaches 90.3% mean success on eight tasks with teams of up to four arms. Ablations show that cross-agent attention and team-stage future-video supervision both improve coordinated control: without attention, a four-arm relay falls from 95.0% to 39.0%, and without this supervision, Place Food falls from 78.0% to 47.0%.
作者:Xianzhe Fan, Chi-Pin Huang, Fu-En Yang, Jing Wang, Yecheng Wu, Yu-Cheng Chou, Johan Bjorck, Linxi "Jim" Fan, Jan Kautz, Yu-Chiang Frank Wang, Zhiding Yu, Hengshuang Zhao, Zhiqi Li
外部来源
本页事实对应的公开页面,以原文为准。(点击将离开本站)