CAREER: Advancing Physical Human-Robot Interaction through Intuitive Sensorimotor Communication

点击翻译按钮获取中文摘要

In the near future, robotic systems will collaborate closely with human partners to complete shared tasks such as materials handling in manufacturing and patient care in clinical settings. To ensure safe and effective interaction, the robotic partner must interpret and understand the human partner's goals and intent, as communicated through various information channels, including physical interaction. The human partner must also be able to interpret and understand the robot's intended actions. This Faculty Early Career Development (CAREER) project aims to understand how a human and a robot can communicate each other’s intent through impedance modulation at a single point of physical contact. The project will use this knowledge to develop a robotic system that interacts safely and intuitively with a human partner as they walk "hand-in-hand" in a collaborative path-following scenario (assisted walking). By analyzing the forces at the hands and consequent movements of the arms during different leader/follower scenarios, specific impedance modulation strategies of humans will be identified that enable natural human-robot interaction through their arms and hands. The project will advance the NSF mission to promote the progress of science and to advance national health, prosperity, and welfare by advancing a fundamental understanding of novel communication strategies made possible through the physical coupling of human and embodied machine intelligences. This project will also provide early exposure and training opportunities on physically interactive robots to the public, particularly to students in STEM programs and to K-12 educators.This work focuses on intent communication and the role of the mechanical impedance of the human arm during collaborative, physical human-robot interactions. The project will measure human arm impedance during guided partner walking with the human acting either as leader or follower. In partner walking, one agent guides the other through a desired path at some desired speed. The robot will apply small transient force perturbations occasionally to the human arm and monitor the resulting arm motions to infer limb impedance. The project team will then identify the contextual modulation of arm impedance during situations where the human assumes the role of the leader and in situations where the human is the follower. By doing so, the project team seeks to discern physical communication strategies underlying effective and intuitive physical human-robot collaborations. Finally, the project will quantify the effectiveness of the identified intent communication strategies by implementing them on a mobile robot that simulates human partnered walking. The experiments will be performed using a force-controlled over-ground interactive robot that walks together with a human in a large room monitored by a 3D motion capture system. The research and educational work of this project will lay the foundation for implementing intuitive physical communication between humans and robots.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.