Ensemble : fluency and embodiment for robots acting with humans

Abstract

This thesis is concerned with the notion of fluency in human-robot interaction (HRI), exploring cognitive mechanisms for robotic agents that would enable them to overcome the stop-and-go rigidity present in much of HRI to date. We define fluency as the ethereal yet manifest quality existent when two agents perform together at high level of coordination and adaptation, in particular when they are well-accustomed to the task and to each other. Based on mounting psychological and neurological evidence, we argue that one of the keys to this goal is the adaptation of an embodied approach to robot cognition. We show how central ideas from this psychological school are applicable to robot cognition and present a cognitive architecture making use of perceptual symbols, simulation, and perception-action networks. In addition, we demonstrate that anticipation of perceptual input, and in particular of the actions of others, are an important ingredient of fluent joint action. To that end, we show results from an experiment studying the effects of anticipatory action on fluency and teamwork, and use these results to suggest benchmark metrics for fluency. We also show the relationship between anticipatory action and a simulator approach to perception, through a comparative human subject study of an implemented cognitive architecture on the robot AUR, a robotic desk lamp, designed for this thesis. A result of this work is modeling the effect of practice on human-robot joint action, arguing that mechanisms that govern the passage of cognitive capabilities from a deliberate yet slower system to a faster, sub-intentional, and more rigid one, are crucial to fluent joint action in well-rehearsed ensembles. Theatrical acting theory serves as an inspiration for this work, as we argue that lessons from acting method can be applied to human-robot interaction.