Exposing real-world challenges using HRI in the wild

We describe an experiment designed to determine actions a robot can perform to reliably enlist help from passers-by in a public space. Based on our experiences with this experiment, we argue in favor of carrying out Human Robot Interaction (HRI) studies in the wild versus in the lab. Our position is that field studies expose challenges that a robot will face that would be difficult to anticipate and simulate in a laboratory setting, and thus will produce results that better generalize to real-world situations. ACM Classification