[论文解读] Humanoid Locomotion and Manipulation: Current Progress and Challenges in Control, Planning, and Learning

Zhaoyuan Gu, Junheng Li+16 more|arXiv (Cornell University)|Jan 3, 2025

Robotic Locomotion and Control被引用 7

一句话总结

本综述综述了人形体运动-操作的进展与挑战，涵盖基于模型的规划/控制、基于学习的方法、基础模型以及全身触觉感知。

ABSTRACT

Humanoid robots hold great potential to perform various human-level skills, involving unified locomotion and manipulation in real-world settings. Driven by advances in machine learning and the strength of existing model-based approaches, these capabilities have progressed rapidly, but often separately. This survey offers a comprehensive overview of the state-of-the-art in humanoid locomotion and manipulation (HLM), with a focus on control, planning, and learning methods. We first review the model-based methods that have been the backbone of humanoid robotics for the past three decades. We discuss contact planning, motion planning, and whole-body control, highlighting the trade-offs between model fidelity and computational efficiency. Then the focus is shifted to examine emerging learning-based methods, with an emphasis on reinforcement and imitation learning that enhance the robustness and versatility of loco-manipulation skills. Furthermore, we assess the potential of integrating foundation models with humanoid embodiments to enable the development of generalist humanoid agents. This survey also highlights the emerging role of tactile sensing, particularly whole-body tactile feedback, as a crucial modality for handling contact-rich interactions. Finally, we compare the strengths and limitations of model-based and learning-based paradigms from multiple perspectives, such as robustness, computational efficiency, versatility, and generalizability, and suggest potential solutions to existing challenges.

研究动机与目标

• 总结在人形体运动与操作的基于模型的规划与控制的最新进展。
• 评估基于学习的方法（包括 RL 和 IL）在 loco-manipulation 任务中的应用与影响。
• 评估基础模型在人形任务规划与语义理解中的作用。
• 强调全身触觉感知在接触丰富的 loco-manipulation 中的重要性，并讨论未来的发展方向。

提出的方法

• 回顾传统的规划与控制方法（接触规划、运动规划、全身控制）。
• 调查应用于 loco-manipulation 的强化学习与模仿学习方法。
• 考察基础模型作为任务规划和语义理解模块的潜力。
• 讨论全身触觉感知作为关键模态，并综合其在手、脚与全身感知中的应用。
• 提供关于基于模型的方法与基于学习的方法整合以实现统一 loco-manipulation 的跨领域讨论。

Figure 1: Humanoids executing locomotion and manipulation tasks: (a) HRP-4 wipes a wood board while adapting to terrain [ 1 ] ; (b-g) Object pick and place by Digit, Hector [ 2 ] , Atlas, H1, Justin [ 3 ] , and Apollo; (h) iCub pushes a cart [ 4 ] ; (i) Nadia opens a door [ 5 ] ; (j-k) Object manipu

实验结果

研究问题

• RQ1 在 humanoid 机器人上实现 loco-manipulation 的最先进的基于模型的方法和基于学习的方法有哪些？
• RQ2 对于多功能且具泛化能力的人形 loco-manipulation 系统，存在哪些尚未填补的空白？
• RQ3 哪些方法在解决这些空白方面有潜力，包括与基础模型和触觉感知的整合？

主要发现

• 基于模型的规划与控制仍然是基础，像 MPC 这样的预测-反应层级和全身控制器是核心。
• 基于学习的方法，特别是 RL 和 IL，正在进步，但在高自由度 humanoids 上面临仿真到现实 transfer 和数据效率的挑战。
• 基础模型提供开放世界推理和语义规划能力，能够支持任务化并统一规划，尽管低级控制仍非端到端。
• 全身触觉感知被日益认为对于接触丰富的 loco-manipulation 与安全至关重要，手、脚和全身感知使新能力成为可能。
• 预计将基于模型的方法与基于学习的策略相结合，以解锁鲁棒、可适应的人形运动与操作。

Figure 2: This survey begins by defining relevant concepts of humanoid robots and their locomotion and manipulation capabilities. Centered around achieving humanoid loco-manipulation tasks, the core of this survey delves into two main categories of methods: the traditional planning and control appro