Methods for control of UGV convoys in unexplored and partially unpaved terrain

Under lab conditions various autonomous behaviors are realized and simulated today, but there is still a long way to everyday applications using these capabilities. Therefore this paper will describe our current research results under competition conditions where, in opposite to presentations or test runs for validation and detailed analysis of algorithms, not everything is planable in advance. We will describe the importance of robustness and intuitive control panels, as well as the “follow the leader„ capabilities of our technology demonstration vehicle, using different levels of autonomy, starting from direct control through teleoperation, finally to fully autonomous terrain-based optimized following. This will bring us to the description of adaptive robot control, where the method for robot control depends on the terrain, the mission, and the availability of supervision by a human operator. Therefore several underlying technologies have been extended or newly developed directly, like an specific A* algorithm, that considers non-holonomic constraints and some basic researches on global optimization methods, like mixed integer programming, and its combination with local reactive collision avoidance methods like VFH+ or Fuzzy Systems.