Modeling Human Behaviour in Emergency: A Research Agenda for the Creation of a Rescue Robot

Abstract

The development of an artificial agent-driven rescue robot capable of interacting naturally with humans under urban emergency conditions brings out a number of issues that need to be addressed: among these are (i) providing a full description and understanding of the user's perspective, and thus creating a reliable and functional user model, (ii) addressing the question of appearance, including the interface to be implemented, and (iii) evaluating the reliability and credibility of such a robot under, both group and individual panic reactions to stressful, extreme conditions. This article addresses these topics with a set of suggestions for future research in the field. In particular, we propose a research agenda focused on the appearance of a rescue agent, and its trustworthiness under extreme circumstances, together with suggestions for a set of experiments aimed at enlighten a user model of crowd and individual behaviour in emergency situations.