arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2610.35862cs.ROcs.AIcs.LGcs.SYeess.SY
收藏

学习力反应型人形机器人行走-操作

Learning Force-Reactive Humanoid Loco-Manipulation

Amazon FAR (Frontier AI & Robotics)(亚马逊前沿人工智能与机器人研究院);New York University(纽约大学);Carnegie Mellon University(卡内基梅隆大学);University of Chicago(芝加哥大学)

Sihang Li, Xiaofeng Guo, Jinkun Cao, Charlie Cheng, Ruihan Yang, Chen Feng, Guanya Shi, Haozhi Qi

AI总结:

提出FLORE框架,结合力反应型扩散策略与混合顺应性全身控制器,在人形机器人行走-操作任务中,力反应型策略成功率80%,远高于纯位置策略的26.7%。

摘要:

许多日常任务需要同时控制接触力和运动。在人形机器人上,这种力通常需要在双手和全身之间协调,无法仅通过运动学轨迹指定。因此,我们提出FLORE,一个用于学习力反应型人形机器人行走-操作的框架。我们使用两个力传感UMI风格手持夹具收集无机器人演示,用于手部运动和接触 wrench,以及可穿戴追踪器用于身体运动。力反应型扩散策略以视觉、本体感觉和测量的wrench作为输入,然后预测动作和每只手的力命令。带有力跟踪的混合顺应性全身控制器执行这些命令同时保持平衡:由真实力/扭矩(F/T)传感器测量的力误差积分使每只手沿命令方向准确跟踪力,并在其他方向保持顺应性。在仿真中,该控制器跟踪不同大小和方向的每只手力命令,全身运动精度损失很小。在机器人上的三个接触丰富任务中,力反应型策略的成功率为80%,而在相同演示上训练并由同一控制器执行的纯位置策略成功率为26.7%。更多细节请访问我们的网站:this https URL。

英文摘要:

Many everyday tasks require controlling contact force as well as motion. On a humanoid, this force must often be coordinated across both hands and the whole body, and cannot be specified by kinematic trajectories alone. Therefore, we present FLORE, a framework for learning force-reactive humanoid loco-manipulation. We collect robot-free demonstrations using two force-sensing UMI-style handheld grippers for hand motion and contact wrenches, and wearable trackers for body motion. A force-reactive diffusion policy takes vision, proprioception, and measured wrenches as input, then predicts actions and per-hand force commands. A hybrid compliance whole-body controller with force tracking executes these commands while maintaining balance: an integral of the force error measured by real force/torque (F/T) sensors lets each hand track force accurately along the commanded direction and remain compliant along the others. In simulation, the controller tracks per-hand force commands of varied magnitude and direction, with little loss in whole-body motion accuracy. Across three contact-rich tasks on the robot, the force-reactive policy succeeds in 80% of rollouts, compared with 26.7% for a position-only policy trained on the same demonstrations and executed by the same controller. For more details, please visit our website: https://louis-leee.github.io/FLoRe/.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

相关深度报道

↑