Dynamic Shared Control for Soft Robots

该奖项支持人机协作研究，同时控制由柔性或可拉伸材料制成的软机器人。软体机器人非常适合现实世界的任务，与传统机器人相比，它们可以更安全、更稳健地与人类互动。然而，预测人类的行动和操作反应对机器人来说可能是具有挑战性的，而人类直接控制机器人可能反过来需要对机器人的运动有直观的理解，这增加了用户的工作量。研究小组将探索共享控制的潜力，这可以帮助软机器人从人类对环境相互作用的洞察中受益，同时也减轻了人类用户的负担。这项研究的发现将有助于推动软机器人融入一系列人类环境，从个人到工业，并可能导致新的紧急行为，提高我们对这些机器人的自主控制。该项目还将接触当地社区和高中生，以提高对STEM学科和人机交互的参与和认识。这项研究将促进对软机器人系统的人类远程操作的基本理解，以及当以任务响应方式共享控制时出现的紧急行为。首先，将研究具有复杂被动和主动运动的系统操作中用户工作量的实时特征，从而更好地理解工具（机器人）特征如何影响操作员的工作量。这将为使用新型机器人时，人类对软机器人的印象与工作量之间的关系提供首次测量。其次，软机器人的力学模型限制了对屈曲和材料失效事件的预测，将为自主和远程操作系统提供更安全的操作。第三，该项目将研究限制的沟通如何影响人类操作员的行为和软机器人的性能。最后，将开发用于软机器人的滑动共享控制算法，从而提高系统的性能，并更好地作为人类操作员的工具。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This award supports research on human-robot collaboration while controlling soft robots made of flexible or stretchable materials. Soft robots are well-suited for real-world tasks and can interact with humans more safely and robustly than traditional robots. However, predicting humans' actions and reactions to operate can be challenging for the robot, while direct control of the robot by humans may, in turn, require an intuitive understanding of the robot's movements, increasing the workload for the user. The research team will explore the potential of shared control, which can help soft robots benefit from human insights into environmental interactions, while also reducing the burden on the human user. The findings of this research will be instrumental in advancing the integration of soft robots into a range of human environments, from personal to industrial, and could lead to new emergent behaviors that improve our autonomous control of these robots. The project will also reach out to the local community and high school students to increase engagement and awareness of STEM disciplines and human-robot interaction. This research will advance fundamental understanding of human teleoperation of soft robotic systems and the emergent behaviors that arise when control is shared in task-responsive ways. First, real-time features of user workload in the operation of systems with complex passive and active movements will be investigated, giving a better understanding of how tool (robot) features affect the operator's workload. This will provide a first-ever measure of the relationship between the human impression of soft robots and the workload when using a novel robot. Second, mechanical models of soft robot limits to predict buckling and material failure events, providing safer operations for autonomous and teleoperated systems will be developed. Third, the project will investigate how communication of limits affects human operator behavior and soft robot performance. Lastly, sliding shared control algorithms for soft robots will be developed, leading to increased performance of the systems and better adoption as tools for human operators.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.