Robot behavior toolkit: generating effective social behaviors for robots

  • Authors:
  • Chien-Ming Huang;Bilge Mutlu

  • Affiliations:
  • University of Wisconsin-Madison, Madison, WI, USA;University of Wisconsin-Madison, Madison, WI, USA

  • Venue:
  • HRI '12 Proceedings of the seventh annual ACM/IEEE international conference on Human-Robot Interaction
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Social interaction involves a large number of patterned behaviors that people employ to achieve particular communicative goals. To achieve fluent and effective humanlike communication, robots must seamlessly integrate the necessary social behaviors for a given interaction context. However, very little is known about how robots might be equipped with a collection of such behaviors and how they might employ these behaviors in social interaction. In this paper, we propose a framework that guides the generation of social behavior for humanlike robots by systematically using specifications of social behavior from the social sciences and contextualizing these specifications in an Activity-Theory-based interaction model. We present the Robot Behavior Toolkit, an open-source implementation of this framework as a Robot Operating System (ROS) module and a community-based repository for behavioral specifications, and an evaluation of the effectiveness of the Toolkit in using these specifications to generate social behavior in a human-robot interaction study, focusing particularly on gaze behavior. The results show that specifications from this knowledge base enabled the Toolkit to achieve positive social, cognitive, and task outcomes, such as improved information recall, collaborative work, and perceptions of the robot.