PACT: Posture-Aligned Co-Design Technique for Hand Configuration and Retargeting in Multi-Finger Imitation Learning
作者: Koyo Sasazaki, Satoshi Funabashi, Pranav Ponnivalavan, Alexander Schmitz, Tetsuya Ogata, Shigeki Sugano
分类: cs.RO
发布日期: 2026-10-13
备注: 9 pages, 6 figures, 2 tables. Submitted to IEEE ICRA 2027. Project page: https://pact-hand-2027.github.io
💡 一句话要点
提出PACT以优化多指模仿学习中的手部配置与重定向
🎯 匹配领域: 支柱一:机器人控制 (Robot Control) 支柱二:RL算法与架构 (RL & Architecture) 支柱七:动作重定向 (Motion Retargeting)
关键词: 模仿学习 机器人手 姿态重定向 机械优化 人机交互
📋 核心要点
- 现有的机器人手部设计往往忽视人类的抓握姿势,导致模仿学习的效果不佳。
- PACT通过联合优化机械参数和姿态重定向器,确保机器人手能够更好地模拟人类手部动作。
- 实验表明,PACT手在抓取精度和成功率上均显著优于传统基线,展示了其在实际应用中的潜力。
📝 摘要(中文)
本研究提出PACT(姿态对齐共设计技术),旨在提高机器人手部在模仿学习中的表现。PACT通过优化17个伺服电机的机械参数和非线性姿态重定向器,确保机器人手能够准确复制人类的抓握姿势。实验结果显示,PACT手在关键点重现上具有3.9毫米的运动学均方根误差,相较于基线LEAP手的12.8毫米有显著提升。此外,PACT手在抓取日常物品时的成功率达到86%,而基线仅为64%。
🔬 方法详解
问题定义:本研究旨在解决现有机器人手部在模仿学习中无法有效复制人类抓握姿势的问题。传统方法通常在硬件设计完成后才选择姿态重定向法,导致抓握效果不理想。
核心思路:PACT的核心思路是基于人类抓握姿势的优化,联合调整机器人手的机械参数与姿态重定向器,以实现更高的姿态重现精度。
技术框架:PACT的整体架构包括机械参数优化模块(如连杆长度、关节偏移、安装角度等)和姿态重定向模块(如人机关节映射、每关节姿态权重和阻尼),并在优化过程中考虑伺服干扰、组装间隙等实际因素。
关键创新:PACT的主要创新在于其联合优化的设计方法,确保了机器人手在抓握时能够实现全手指对应的人类姿态,而不是单独优化某一部分。
关键设计:在设计中,PACT手的伺服电机参数经过精确调整,损失函数考虑了姿态重定向的准确性和机械可组装性,确保最终设计能够在物理上实现。实验中,PACT手的运动学均方根误差为3.9毫米,显著优于基线的12.8毫米。
🖼️ 关键图片
📊 实验亮点
PACT手在实验中表现出色,关键点重现的运动学均方根误差为3.9毫米,远低于LEAP手的12.8毫米。此外,PACT手在抓取日常物品时的成功率达到86%,而基线仅为64%,显示出显著的性能提升。
🎯 应用场景
PACT的研究成果在机器人手部设计、远程操作和人机交互等领域具有广泛的应用潜力。通过提高机器人手的抓握精度和成功率,PACT能够在医疗、服务机器人以及工业自动化等场景中发挥重要作用,推动智能机器人技术的发展。
📄 摘要(原文)
The fidelity of teleoperated demonstrations used for imitation learning is determined by how faithfully a robot hand can replicate the operator's posture. The posture retargeting law that the robot will execute is typically selected only once the hardware has been finalized, and conventional hand designs often neglect human-like grasping postures. We introduce PACT, an optimization based on human grasping postures that jointly optimizes the mechanical parameters of a 17-servo robot hand (link lengths, joint offsets, mounting angles, and base placements) and its nonlinear posture retargeter (a linear human-to-robot joint map, per-joint posture-IK weights, and damping). The optimization enforces full-finger correspondence (MCP, PIP, DIP, and tip of each finger) to human poses from the DexYCB dataset, and is constrained by servo interference, assembly clearances, and flat-hand protrusion, so the resulting design can be physically assembled. We fabricate the PACT Hand and validate it through teleoperation experiments and an imitation-learning demonstration. The hand reproduces all human hand keypoints with a kinematic RMSE of 3.9 mm, compared to 12.8 mm for the LEAP Hand baseline, and keeps all fingertips within 10 mm on every grasping posture, while the baseline achieves this for only 21.6% of the poses. Under matched per-joint torque limits, it lifts twice the baseline payload, 2 kg against 1 kg, and deforms a held paper cup seven times less, reaching a deformation level comparable to a human hand. It also achieves an 86% grasping success rate on YCB daily objects, compared to 64% for the baseline. Finally, as a demonstration, an imitation-learning policy trained on demonstrations collected with the PACT Hand achieved a 75% success rate at held-out placements in a can pick-and-place task. Project page: https://pact-hand-2027.github.io