Path planning for off-road vehicles with a simulator-in-the-loop

This paper describes the development of a real-time path planner for off-road vehicles using a simulator. The work was triggered by a need for an obstacle-avoidance and path-planning system in our work with autonomous forest machines. The general idea with the presented system is to extend a standard path-tracking algorithm with a simulator that, in real-time, tries to predict collisions in a window forward in time. This simulation is based on current sensor data giving information about the environment around the vehicle. If a collision is predicted, the vehicle is stopped and a path-search phase is initiated. Variants of the original path are generated and simulated until a feasible path is found. The real vehicle then continues, now tracking the replanned path. In simulated tests, this way of using a simulator to predict and avoid collisions works well. The system is able to safely navigate around obstacles on and close to the path in a way that is hard or impossible to achieve with standard obstacle-avoidance algorithms that do not take the shape of the vehicle into account. Another scenario, also envisioned in forest environment, is off-line path planning of a longer route, based on map information. An approximate path given by a straight line from start to goal is then modified in the same way as described above