Autonomous Social Robots in Pedestrian-Rich Environments

Robots operating in human environments have to manage many difficult challenges, including navigating in open, dynamic environments as well as direct and indirect interactions with surrounding pedestrians. Consider the scenario of walking down the sidewalk to deliver a package to a customer. The robot needs to be able to localise itself in an open environment, it needs to be able to recognise sidewalk boundaries, traffic lights, and crosswalks, it needs to be able to move along the sidewalk safely while sharing space with pedestrians, pets, cyclists, and it needs to be able to understand and follow certain social/cultural conventions and expectations, which are often ill-defined, in order to be accepted by the pedestrians whom it shares the sidewalk with.

This project focuses on developing the technical and social capabilities an autonomous robot needs for operating in shared spaces with humans. In collaboration with the University of British Columbia, we have been working on developing robot skills such as person following for outdoor navigation, user recognition and tracking for personalised interaction, as well as non-verbal communicative displays for conveying robot intent. Target applications include homecare robots, assisted living facility staff assistants, and last-mile delivery robots.