Blind Precise Assembly: Proprioceptive Sim-to-Real Peg-in-Hole for Diverse Parts
作者: Antonia Bronars, Younghyo Park, Kushal Kedia, Pulkit Agrawal
分类: cs.RO
发布日期: 2026-10-16
备注: 21 pages, 15 figures, 7 tables. Antonia Bronars and Younghyo Park contributed equally. Project page: https://younghyopark.me/blind-proprioceptive-assembly/
💡 一句话要点
提出盲目精确装配方法以解决复杂部件组装问题
🎯 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)
关键词: 机器人装配 本体感知 强化学习 盲装配策略 多样化部件 模拟训练 泛化能力
📋 核心要点
- 现有的机器人装配方法依赖于视觉和触觉传感器，难以处理复杂和多样化的部件组装任务。
- 本研究提出了一种盲装配策略，利用机器人自身的关节传感器进行精确的接触感知，避免了对外部传感器的依赖。
- 实验结果表明，盲装配策略在真实环境中成功插入不同形状和尺寸的插销，且在未见部件上具有更好的泛化能力。
📝 摘要（中文）
本研究探讨了在没有视觉、力、扭矩或触觉传感的情况下，机器人如何通过自身关节的本体感知进行精确的接触丰富装配。我们训练了盲装配策略，仅依赖关节位置、速度和自身动作的短期历史，面对随机化的夹具和部件。通过强化学习在模拟环境中训练，这些策略成功实现了插销插入、齿轮啮合和螺母穿线等任务。插销插入策略在真实世界中实现了73%的成功率，且与视觉策略的对比显示，盲策略在未见部件形状和尺寸上具有更好的泛化能力。
🔬 方法详解
问题定义：本研究旨在解决机器人在复杂部件组装中对视觉和触觉传感器的依赖，现有方法在处理微米级接触时存在局限性。
核心思路：提出了一种盲装配策略，通过机器人关节的本体感知进行装配，完全依赖于关节位置和速度的历史数据，而不使用任何外部传感器。
技术框架：整体流程包括在模拟环境中使用强化学习训练盲装配策略，策略通过随机化的部件和夹具进行训练，学习插销插入、齿轮啮合和螺母穿线等任务。
关键创新：最重要的创新在于通过本体感知实现了精确装配，显著提高了对未见部件形状和尺寸的泛化能力，与传统依赖视觉的策略相比，具有更强的适应性。
关键设计：训练过程中使用了大量不同的装配实例，确保每个实例的间隙保持完整，采用了开放源代码的生成器，能够从精确的凸形状构建装配，提升了模拟的准确性和速度。
🖼️ 关键图片
📊 实验亮点
实验结果显示，盲装配策略在真实世界中成功插入不同形状和尺寸的插销，成功率达到73%。与视觉策略的对比表明，盲策略在未见部件上具有更好的泛化能力，尽管速度可能较慢，但在多样化部件上表现更为出色。
🎯 应用场景
该研究的潜在应用领域包括自动化装配线、机器人制造和服务机器人等。通过减少对外部传感器的依赖，机器人能够在复杂环境中更灵活地进行装配，提高生产效率和适应性。未来，该方法可能推动机器人在更广泛的工业和日常生活场景中的应用。
📄 摘要（原文）
Every robot already carries high-resolution contact sensors: its own joints. Legged robots and humanoids exploit them to walk, run, and track whole-body motions without vision, but there the contacts are large and forceful. Is joint sensing also enough for precise, contact-rich assembly, where the robot must find a millimeter-scale opening, mesh gear teeth, or catch a thread through light contact? We study this question by training blind assembly policies that observe only a short history of joint positions, velocities, and their own actions, without visual, force, torque, or tactile sensing. The poses of both the fixture and the part are unknown to the policy, and the grasp is randomized substantially. The policy must therefore assemble the parts through contact, sensed exclusively through its own proprioception. Trained with reinforcement learning entirely in simulation over randomized parts, grasps, and contacts, such policies learn peg insertion, gear meshing, and nut threading. The peg insertion policy transfers zero-shot to the real world and inserts pegs of different shapes and sizes in 73% of trials. Controlled comparisons with matched visual policies in simulation show that blind policies generalize far better to unseen part shapes and sizes, though they can be slower. Training such policies needs thousands of different assemblies, each simulated with its clearance intact, so we also release an open-source generator that builds them procedurally from exact convex pieces for MuJoCo; it is more accurate and faster than convex decomposition or MuJoCo's signed-distance collision. Videos and interactive simulations are at https://younghyopark.me/blind-proprioceptive-assembly/.