Mobile robot navigation in unknown dynamic environment inspired by human pedestrian behavior

Abstract

Navigation in an unknown dynamic environment is one of the key challenges in mobile robotics. This paper proposes a scheme, inspired by human pedestrian behavior, for navigation of a mobile robot in an a priori unknown dynamic environment. An occupancy grid map has been built using onboard sonar sensors through successive sensor information. Inspired by human pedestrian behavior to maintain a safe direction and distance to avoid collisions with obstacles, the proposed navigation scheme trail a path for the robot following a forbidden region map concept with a velocity proportional to the distance and rate at which the obstacles are approaching or receding the robot. The reachable region of robot navigation horizon is based on the motion model predictability of the obstacles. The navigation scheme is deployed on a Fire Bird V mobile robot. The experimental result shows that the robot is able to follow a smooth and time-efficient path avoiding collisions with the mobile and stationary obstacles.