NRI: Collaborative Research: Cooperative Control of Humanoid Robots for Remote Operations in Nuclear Environments

随着使能技术的成熟，类人机器人将获得大量的实际应用，如家庭中的个人助理、救灾中的第一反应者，以及预先部署的空间探索资产。该项目将推进NASA的人形机器人Valkyrie的能力，以取代人工手，在核设施中常见的现有科学手套箱中进行安全和规避风险的操作。该项目的动机是这样一个事实，即对于人类监督的机器人来说，肮脏、枯燥和危险的任务在当今使用的数千个手套盒中广泛存在。仿人机器人的风险规避控制框架将为机器人学和机器人操作领域做出重大贡献；由此产生的技术还将显著提高人类的安全性、操作的准确性和与高后果材料处理应用中的操作相关的成本效益。该项目的成功完成不仅将使人形机器人在实际应用中做好技术准备，还将为该领域提供新的知识，包括如何将人类的命令转化为安全的机器人控制方法。该项目将创建计划和工具，以培训具有包括远程操作和界面设计在内的新技能的劳动力。该项目的研究目标将为受限机器人操作和人机交互领域做出新的贡献。具体地说，这个项目将(1)验证NASA的人形机器人Valkyrie在人类监督下的控制框架，以避免风险和直观地操作，执行科学手套箱中的常见任务，如将物体推到卸货口；(2)开发基于感知的机器人操作的理论和实践，并设计新的解决方案，以处理不确定性，将运动/力规划与高水平的人类输入和环境智能相结合，以强有力地执行实际任务；(3)介绍和改进操作原理，并评估在核操作中部署人形机器人的安全意义；(4)通过在逐项任务的基础上优化性能，为类人机器人在核清理中的适用性开发新的范例。项目成果将在硬件/软件协同设计、控制理论、机器人学和人-机器人交互的交叉领域创造新的方法来应对多学科挑战。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

As the enabling technologies mature, humanoid robots will find numerous practical applications, as personal assistants in the home, first-responders in disaster relief, and as pre-deployed space exploration assets. This project will advance the capabilities of NASA's humanoid robot, Valkyrie, to replace human hands for safe and risk-averse operations in existing science gloveboxes commonly found in nuclear facilities. The project is motivated by the fact that the dirty, dull and dangerous tasks ideal for human-supervised robots are extensively found in thousands of gloveboxes being used today. The risk-averse control framework for humanoid robots will provide significant contributions to robotics and robot manipulation fields; the resulting technologies will also significantly improve the safety of humans, accuracy of operations, and cost-effectiveness associated with operations in high-consequence material handling applications. Successful completion of this project will not only progress the technological readiness of humanoids for practical applications but also will contribute new knowledge to the field including how to translate human commands to safe robot control methods. The project will create programs and tools to train a workforce with new skills including teleoperation and interface design. The research goals of the project will make novel contributions to the fields of constrained robot manipulation and human-robot interaction. Specifically, this project will (1) validate a human-supervised control framework for the risk-averse and intuitive operation of NASA's humanoid robot, Valkyrie, performing common tasks found in science gloveboxes such as pushing and sliding objects to the discharge ports; (2) develop theory and practice for perception based robot manipulation and design novel solutions to handle uncertainty, blend motion/force planning with high-level human input and ambient intelligence to robustly execute practical tasks; (3) introduce and refine the operational principles and evaluate safety significance of deploying humanoids in nuclear operations; (4) develop new paradigms for applicability of humanoid robots in nuclear clean-up by optimizing performance on a task-by-task basis. The project outcomes will create new approaches in addressing multidisciplinary challenges at the intersection of hardware/software co-design, control theory, robotics, and human-robot interaction.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.