Towards autonomously traversing complex obstacles with mobile robots with adjustable chassis

Looking at the mobility of robots and their chassis most are limited to fairly flat environments. In urban environments common structures such as steps or stairs pose invincible obstacles for such systems. When it comes to unstructured outdoor environments a vast variety of obstacles is imaginable which are not traversable by common robots, for instance boulders, debris, rocks or trunks of fallen or chopped trees. However, there are mobile robots with adjustable chassis providing a higher degree of mobility and enabling them to overcome such obstacles. This paper presents first results on our motion planning algorithm which aims at utilizing the enhanced capabilities of those robots. It takes into account the chassis configuration and the system stability to propose the best path. We use a high level planner to quickly generate a preliminary path by considering the platform's operating limits. We then distinguish between path segments on flat and rough terrain. For each hard segment we restrict the search space to a tube around the initial path. A subsequent planner is used to refine the preliminary path by considering the actuator positions, the robot's stability and a ground contact factor. Our planning algorithm is general in the sense that we do not categorize obstacles and do not use predefined motion sequences for those obstacle classes. Finally, we present a discussion including an analysis of the time complexity and a simulation application in ROS and Gazebo as proof of feasability.