Designing Algorithms For Socially Competent Robotic Navigation.

Despite the great progress in robotic navigation in the past decades, navigating a human environment remains a hard task for a robot, due to the lack of formal rules guiding traffic, the lack of explicit communication among agents and the unpredictability of human behavior. Inspired by the efficiency of human navigation, we employ the insights of sociology studies on pedestrian behavior and psychology studies on action interpretation to design an online planning framework that leverages the power of implicit communication to generate legible robot behaviors in pedestrian environments. The foundation of our approach is a novel topological representation, based on braid groups. Preliminary results demonstrate the efficiency of our approach in simulation, whereas planned experiments with human subjects are expected to enable us to extract realistic predictive models and get user feedback. Finally, we plan on evaluating our approach by running our algorithms on our social robot in crowded environments.