A Microsociological Approach to Understanding the Robot Collaborative Motion in Human-Robot Interaction

Abstract

Existing approaches to human-robot collaboration typically focus on how to build robots that can work safely and fluently with humans on collaborative tasks. Less is known about how people interpret the boundary between movement-based collaboration and non-collaboration when interacting with robots. By applying a microsociological theory to analysing the process of interaction as it unfolds, we propose and identify points of breakdown in the collaborative task of a human and a robot carrying an object to a destination. In designing the experiment, Kinetography Laban is used to enable a precise description of the intended normative and disruptive motion paths of the robot. The contribution of the paper is the theoretical understanding of collaboration from sociology, and a method for designing and evaluating collaborative motion between humans and robots that combines microsociology and Kinetography Laban. The proposed method accounts for the contingent meaning construction performed by people in recognising behavioural motion cues of robots as part of an ongoing interaction process, and enables the boundary between collaborative and non-collaborative robot motion to be defined.