Multi-party Conversation of Driving Agents

Abstract

Recently, the applications of conversational robots have been gaining popularity due to their potential in providing information while engaging the user in a conversation. However, when a user's attention is already focused on a task, engaging them in conversation may be difficult or even risky. The Human-Robot Interaction (HRI) field should consider interaction methods where a conversational robot can keep the user informed but without the obligation of engagement in the conversation. In this study, we discuss such an approach within a driving scenario by utilizing a multi-party social robot platform that comprises three minimalistic conversational robots, which possesses the feature of being able to decrease the number of directed utterances toward a driver through a turn-taking process among the robots. The results of this study revealed that overhearing information from the multi-party conversation of driving agents is perceived as possessing more life-like characteristics compared to a conventional, one-to-one communication-based approach that directly addresses the driver. Moreover, the proposed approach reduced the distraction level and increased the enjoyment of the drivers.