Obstacle-free path planning for mobile robots

In the path planning phase, the representation of the workspace obtained from the preprocessing phase is searched or explored for an obstacle-free path. This search could be for an optimal path, for the most efficiently obtainable path, or for just an obstacle-free path between a given source destination pair. The paper introduces a new general algorithm for the obstacle-free path planning for mobile robots. The proposed technique is based on a simple method of exploring a partially known or fully known work environments. The concept of recoding the image with a quadtree, and only looking at image details when one needs to is sound; the time savings is substantial over the preprocessing methods. The one drawback of this technique is that in the worst case scenario, the planner would not find a path, but it would never find a path that would cause a collision, and the worst case only occurs when the dynamic object expansion is being used.<
>