Socially Inspired Motion Planning for Mobile Robots in Populated Environments

Nowadays, robots often operate in environments that they share with humans. The ability to act similar to humans is an important prerequisite for the social acceptance of robots. In this paper, we consider the problem of navigation in populated environments. We present a path planning algorithm that enables robots to move efficiently and smoothly with groups of people by selecting those individuals that move towards the robot’s desired goal. Our technique is based on a people tracking system in combination with an iterative A* planner. The approach iteratively finds both, a path and a partition of the set of surrounding people into obstacles and subjects to follow. In the absence of people, the optimal solution is still found by the A* planner. The approach has been implemented and tested on a real mobile robot in populated environments. Experiments illustrate that the robot is able to move with groups of people resulting in a more human-like way of navigation among people.