The impact study of uncertainty on robot autonomy range

In the high uncertainty situation, robots perform missions with low plan success rate and a lot of work pressure. In this paper, an approach to build relationship between uncertainty and autonomy is proposed. The essence of the method is that by lowering autonomy range and asking human for heuristic information, the cognitive and decision ability of the system is improved under the uncertainty of situation. The method consists of two parts. First, proposed the possible types of uncertainty are defined evaluation. Second, built the relationship between uncertainty and autonomy. Finally, it proved that the approach is feasible and effective for improving plan success rate by USV navigation application in the simulation system.