RobotSmith：用于获取复杂操作技能的生成式机器人工具设计
RobotSmith: Generative Robotic Tool Design for Acquisition of Complex Manipulation Skills
AI总结：
针对现有机器人工具设计方法的局限，提出RobotSmith自动化流程，结合VLM隐式物理知识与物理仿真，迭代设计工具、生成轨迹并联合优化，在多类操作任务中成功率优于基线，且可迁移到真实场景。
摘要：
赋予机器人工具设计能力，对于让它们能够解决原本难以处理的复杂操作任务至关重要。尽管近期的生成式框架能够自动合成任务设置，例如3D场景和奖励函数，但它们尚未解决工具使用场景的挑战。简单检索人类设计的工具可能并非理想方案，因为许多工具（例如擀面杖）难以被机器人操作器操控。此外，现有的工具设计方法要么依赖预定义模板、仅能进行有限的参数调整，要么采用未针对工具创建优化的通用3D生成方法。为解决这些局限，我们提出RobotSmith，这是一个自动化流程，它利用视觉语言模型（VLM）中蕴含的隐式物理知识，结合物理仿真提供的更精准物理特性，来设计并使用机器人操作所需的工具。我们的系统包含三个核心环节：（1）通过协作式VLM智能体迭代提出工具设计方案；（2）生成工具使用的底层机器人轨迹；（3）针对任务表现联合优化工具几何结构与使用方式。我们在涉及刚性物体、可变形物体和流体物体的广泛操作任务中评估了所提方法。实验表明，我们的方法在任务成功率和整体表现上始终优于强基线方法。值得注意的是，我们的方法实现了50.0%的平均成功率，显著超过其他基线，例如3D生成方法（21.4%）和工具检索方法（11.1%）。最后，我们将系统部署在真实环境中，证明生成的工具及其使用方案能够有效迁移到物理执行过程，验证了我们方法的实用性和泛化能力。
英文摘要：
Endowing robots with tool design abilities is critical for enabling them to solve complex manipulation tasks that would otherwise be intractable. While recent generative frameworks can automatically synthesize task settings, such as 3D scenes and reward functions, they have not yet addressed the challenge of tool-use scenarios. Simply retrieving human-designed tools might not be ideal since many tools (e.g., a rolling pin) are difficult for robotic manipulators to handle. Furthermore, existing tool design approaches either rely on predefined templates with limited parameter tuning or apply generic 3D generation methods that are not optimized for tool creation. To address these limitations, we propose RobotSmith, an automated pipeline that leverages the implicit physical knowledge embedded in vision-language models (VLMs) alongside the more accurate physics provided by physics simulations to design and use tools for robotic manipulation. Our system (1) iteratively proposes tool designs using collaborative VLM agents, (2) generates low-level robot trajectories for tool use, and (3) jointly optimizes tool geometry and usage for task performance. We evaluate our approach across a wide range of manipulation tasks involving rigid, deformable, and fluid objects. Experiments show that our method consistently outperforms strong baselines in terms of both task success rate and overall performance. Notably, our approach achieves a 50.0\% average success rate, significantly surpassing other baselines such as 3D generation (21.4%) and tool retrieval (11.1%). Finally, we deploy our system in real-world settings, demonstrating that the generated tools and their usage plans transfer effectively to physical execution, validating the practicality and generalization capabilities of our approach.
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。
