迈向机器人运动中的具身规模定律
Towards Embodiment Scaling Laws in Robot Locomotion
AI总结：
该研究通过程序化生成约千种差异化机器人具身并训练运动策略，验证了具身规模定律，证明增加训练具身比固定具身扩数据更能提升泛化，且策略可零样本迁移至真实Unitree Go2和H1。
摘要：
跨具身泛化是构建适用于任意机器人的通用具身智能体愿景的基础，但其促成因素仍知之甚少。我们研究具身规模定律，即增加训练具身数量可提升对未见具身泛化能力的假设，并以机器人运动作为试验平台。我们通过程序化方式生成约1,000种具有拓扑、几何和关节级运动学差异的具身，并在随机子集上训练策略。我们观察到支持该假设的正向规模趋势，并发现与在固定具身上进行数据规模扩展相比，具身规模扩展能够实现显著更广泛的泛化。我们在完整数据集上训练的最佳策略可零样本迁移到仿真和真实世界中的新型具身，包括Unitree Go2和H1。这些结果代表着迈向通用具身智能的一步，并与可配置机器人的自适应控制、形态协同设计等领域相关。
英文摘要：
Cross-embodiment generalization underpins the vision of building generalist embodied agents for any robot, yet its enabling factors remain poorly understood. We investigate embodiment scaling laws, the hypothesis that increasing the number of training embodiments improves generalization to unseen ones, using robot locomotion as a test bed. We procedurally generate ~1,000 embodiments with topological, geometric, and joint-level kinematic variations, and train policies on random subsets. We observe positive scaling trends supporting the hypothesis, and find that embodiment scaling enables substantially broader generalization than data scaling on fixed embodiments. Our best policy, trained on the full dataset, transfers zero-shot to novel embodiments in simulation and the real world, including the Unitree Go2 and H1. These results represent a step toward general embodied intelligence, with relevance to adaptive control for configurable robots, morphology co-design, and beyond.
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。
