论文arXiv 2611.189765 天前
TMA框架实现Unitree G1零样本Sim-to-Real触觉驱动强力操作
本文由 AI 根据公开内容整理,可能有误,请以原文为准:TMA: Sim-to-Real Forceful Humanoid Manipulation via Tactile-Driven Motor Adaptation(arXiv 2611.18976)
核心内容
针对人形机器人在非结构化环境中多指强力操作的接触不确定性挑战,研究团队提出触觉驱动运动适应(TMA)框架及全身控制系统。该系统以旋转门把手开启为基准任务,要求机器人具备稳定的多指摩擦抓取、防滑扭矩生成及力控推压能力。实验验证表明,TMA使Unitree G1人形机器人配合BrainCo Revo2多指灵巧手,在四种真实商用圆门把手上实现了零样本Sim-to-Real迁移,操作成功率大幅超越标准快速运动适应(RMA)及反应式触觉策略基线。
TMA核心机制是将策略学习锚定于触觉反馈可观测的交互特征,引入锥体几何作为多指接触表征的归纳偏置,并基于触觉历史信息在线自适应调整电机行为。系统架构采用分层设计,上层操作策略与预训练通用步态控制器协同,兼顾灵巧交互与全身稳定性。整个训练过程完全在仿真中完成,无需真实世界微调即可部署至物理本体,有效解决了强力操作中接触动力学难以精确建模导致的Sim-to-Real鸿沟问题。
关键看点
- 触觉适应机制:TMA利用锥体几何归纳偏置表征多指接触,基于触觉历史在线调整电机行为,解决强力操作中的接触不确定性。
- 零样本迁移验证:Unitree G1搭载BrainCo Revo2灵巧手,在四种商用圆门把手上实现纯仿真训练到真机的零样本部署。
- 全身协同控制:上层操作策略与预训练通用步态控制器并行运行,确保多指摩擦抓取与力控推压时的全身动态稳定性。
- 性能显著领先:在真实门把手开启任务中,TMA的成功率与鲁棒性均大幅优于标准RMA算法及传统反应式触觉策略基线。
工程与产业观察
TMA证明了将触觉反馈作为Sim-to-Real核心锚点的有效性,其锥体几何接触表征与在线电机适应机制为人形机器人灵巧操作提供了新范式。对于算法工程师而言,该工作表明在缺乏精确物理参数时,利用触觉归纳偏置比单纯依赖视觉或本体感知更能弥合仿真与现实的动力学鸿沟,对具身智能在非结构化场景下的落地具有直接参考价值。
英文摘要 (Abstract)
Recent humanoid loco-manipulation in articulated environments has largely focused on geometry-dominant interactions, where success is primarily determined by geometric alignment and motion execution. In contrast, forceful multi-finger manipulation requires the robot to adapt to real-world uncertainties in contact and interaction dynamics. We present Tactile-Driven Motor Adaptation (TMA) and an integrated whole-body system for zero-shot sim-to-real forceful humanoid manipulation, and study round-doorknob opening as a challenging benchmark requiring stable multi-finger frictional grasping, sufficient torque generation without slip, force-regulated pushing, and whole-body stability. TMA structures policy learning around interactions observable through tactile feedback, uses cone geometry as an inductive bias for multi-finger contact representation, and adapts motor behavior online from tactile history. An upper-body manipulation policy is trained alongside a pretrained general locomotion controller to coordinate dexterous interaction and whole-body stabilization. Trained entirely in simulation, TMA transfers zero-shot to a Unitree G1 humanoid with BrainCo Revo2 multi-fingered hands and substantially outperforms standard Rapid Motor Adaptation (RMA) and reactive tactile policy baselines across four commercial round doorknobs in the real world.
作者:Yiting Chen, Kelin Yu, Sarthak Kaingade, Bryan Trinh, Kaiyu Hang, Roberto Martin-martin, Nima Fazeli, X. Alice Wu
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)