Complete coverage path planning of mobile robots for humanitarian demining

Purpose - The purpose of this paper is to present a path planning algorithm for a non-circular shaped mobile robot to autonomously navigate in an unknown area for humanitarian demining. For that purpose the path planning problem comes down to planning a path from some starting location to a final location in an area so that the robot covers all the reachable positions in the area while following the planned path. Design/methodology/approach - The proposed algorithm uses occupancy grid map representation of the area. Every free cell in the grid map represents a node in the graph being searched to find the complete coverage path. The complete coverage path is followed by the dynamic window algorithm, which includes robot's kinematic and dynamic constraints. Findings - The proposed algorithm finds the complete coverage path in the graph accounting for the dimensions of the mobile robot, where non-circular shaped robots can be easily included. The algorithms are implemented under the ROS (robot operating system) and tested in the stage 3D simulator for mobile robots with a randomly generated simulation map of an unknown area. Research limitations/implications - Some parts of the area close to obstacles are hard to cover due to complex non-circular shaped robot and non-perfect path following. The future work should include better path following algorithm. Practical implications - The proposed algorithm has shown itself as effective and could meet the working demands of humanitarian demining. Originality/value - The algorithm proposed in the paper enables complete coverage path planning of non-circular shaped robots in unknown areas.