Human-aware mobile robot exploration and motion planner

Abstract

Conventionally, robot exploration problem assumes a static environment. Casting the exploration problem with the presence of humans brings new possibilities for interactive exploration tasks including robot rescue, etc. In this paper, we present a mobile robot system designed to explore an unknown environment with humans. The system is able to perform the robot localization and mapping, and produce the positions of humans in the environment at the same time. The system combines techniques of simultaneous localization and mapping (SLAM), motion planning, and human recognition. Our approach maximizes the exploration efficiency by designing optimal exploration paths. Moreover, the robot is able to perform human-aware behavior during the exploration, which complies with human safety, preferences and actions. Both simulation and field experimental results in real world settings are presented.