Airport snow shoveling

In this paper, we present results of a feasibility study of airport snow shoveling with multiple formations of autonomous snowplow robots. The main idea of the approach is to form temporary coalitions of vehicles, whose size depends on the width of the roads to be cleaned. We propose to divide the problem of snow shoveling into the subproblems of task allocation and motion coordination. For the task allocation we designed a multi-agent method applicable in the dynamic environment of airports. The motion coordination part focuses on generating trajectories for the vehicle formations based on the output of the task allocation module. Furthermore, we have developed a novel approach of formation stabilization into variable shapes depending on the width of runways. The method using a receding horizon control provides optimal trajectories and inputs for robots´ actuators during splitting and coupling of formations. The algorithm can be utilized in arbitrary static and dynamic airport assemblage. All components as well as the complete system have been verified in various simulations and hardware experiments in both indoor and outdoor environments, which are presented in the submitted video.