arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2408.05301cs.RO
收藏

与REEM-C共舞:一项机器人对人类的物理-社会通信研究

Dancing with REEM-C: A robot-to-human physical-social communication study

Marie Charbonneau, Francisco Javier Andrade Chavez, Katja Mombaur

AI总结:

本文以leader-follower舞蹈场景为切入点,开发结合导纳、阻抗与位置控制的分体式全身控制框架,通过实验验证控制器性能,探究机器人对人类的清晰直观通信及人类舒适度问题。

摘要:

人类常紧密协作,通过物理交互传递大量信息。而机器人目前尚无法与人类实现类似的紧密协作,也难以在物理-社会人机交互(psHRI)中有效传递信息,这一现状限制了人机协作解决现实世界问题的潜力。本文研究如何在物理-社会人机交互(psHRI)中建立清晰直观的机器人对人类通信,同时兼顾人类的舒适度。我们从 leader-follower 舞蹈场景切入,其中全身人形机器人通过触觉、视觉、听觉等多种通信模态发出信号,引导人类完成下一步动作。该场景的实现依托于分体式全身控制框架的开发,该框架结合了上半身的导纳控制与阻抗控制,以及下半身用于平衡和迈步的位置控制。通过机器人主导的物理-社会人机交互(psHRI)参与者实验,我们验证了控制器性能,同时从人类伙伴的视角深入理解了哪些类型的通信效果更好,尤其是在感知有效性和舒适度方面。

英文摘要:

Humans often work closely together and relay a wealth of information through physical interaction. Robots, on the other hand, are not yet able to work similarly closely with humans and to effectively convey information when engaging in physical-social human-robot interaction (psHRI). This currently limits the potential of human-robot collaboration to solve real-world problems. This paper investigates how to establish clear and intuitive robot-to-human communication, while considering human comfort during psHRI. We approach this question from the perspective of a leader-follower dancing scenario, in which a full-body humanoid robot leads a human by signaling the next steps through a choice of communication modalities including haptic, visual, and audio signals. This is achieved through the development of a split whole-body control framework combining admittance and impedance control on the upper body, with position control on the lower body for balancing and stepping. Robot-led psHRI participant experiments allowed us to verify controller performance, as well as to build an understanding of what types of communication work better from the perspective of human partners, particularly in terms of perceived effectiveness and comfort.

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。

↑