Route Learning and Reproduction in a Tour-Guide Robot

Route learning and reproduction in tour-guide robots is usually performed with the help of an expert in robotics. In this paper we describe a novel approach to these tasks, which reduces the intervention of an expert to a minimum. First, the robot is able to learn routes while following a human acting as a route instructor. Then, anyone can easily ask the robot to reproduce a route using various hand gestures. In order to achieve an accurate route learning and reproduction we use a novel localization algorithm, which is able to combine various sources of information to obtain the robot’s pose. Moreover, the path planning and obstacle avoidance used to navigate while reproducing routes are also described in this article. Finally, we show through several trajectories how the robot is able to learn and reproduce routes.