A Software System for Human-Robot Interaction To Collect Research Data

Abstract

A key area in human-robot interaction research is the use of a robot to collect participant research data. However, traditional website-based data collection methods do not intergrate with the robot providing the interaction. This leaves a clear disconnect between the static delivery of a digital questionnaire and a lack context-relevant behaviours from the robot. In this paper, we present a HTML/Javascript software system to create a direct link between digital data collection and the robot that delivers it. In doing so, this system can be used to create more dynamic data collection sessions using the robot's speech, movement or presence to support the delivery of questionnaires. This system can also be used to create interactive sessions to support experimentation. We present two proposed use-case scenarios built with the system with the Pepper humanoid robot in mental health and well-being services. We present this software system to help speed up development time for user studies, lower the entry barrier for non-technical researchers who want to use social robots for data collection, and to create more systematic data collection methods for robots. Future work of the software includes increasing the repertoire of questionnaire items available to allow for more sophisticated data collection.