Building Blocks for Human Robot Interaction

点击翻译按钮获取中文摘要

The aim of this research is to identify and develop fundamental interaction building blocks needed to permit the rapid and flexible development and deployment of personal and public robotics. The focus of the proposed work is on transitional interactions where cues about changes in collaborative activity are communicated. These cues signal transition related information essential to the collaboration flow such as: turn taking/giving and role changes (e.g., leader/follower, instructor/trainee). This model emulates the natural ability of small human teams to manage interaction transitions via non-verbal cues. The guiding research questions my group will ask and answer are: i) How do we describe and taxonomize the common transitional interactions that permit successful collaborations with robotic agents? ii) What are the bilateral, "handshaking" cues related to these transitions? iii) How can robots and people best recognize and transmit these cues in natural, untrained (interface/semaphore free) collaborations? iv) What control methods allow the robot to successfully participate in the interaction? v) How do we measure the effectiveness of communication and detection of interaction cues - what are our baselines, and can we identify key metrics allow us to compare efficacy with others? and vi) How do we demonstrate and benchmark these human-robot communication methods. This work provides a basis for the natural integration and acceptance of robotic assistants into the human sphere in a wide range of scenarios. My students and I will make novel and fundamental contributions in the fast growing field of human-robot interaction design, which intersects work in human computer interaction, social robotics, mechatronics and controls. In particular, we will contribute to the understanding and design of transitional interactions between humans and robots and ultimately apply our results to spin-off robotics research projects in manufacturing and healthcare.