A Coordination Strategy for Multi-Robot Sampling of Dynamic fields

A coordination mechanism to achieve the sampling task of static or dynamic fields by means of a system composed by multiple mobile robots is addressed in this paper. The problem is the estimation of a scalar field. To this aim in a probabilistic framework a solution is proposed that takes into account several constraints. The attention is focused on the vehicles motion generation and the developed strategy is designed for multiple, autonomous and distributed robots. It makes use of the Voronoi tessellation's properties to automatically distribute the vehicles' motion and of the Null-Space-Behavioral control to handle eventually conflicting motion tasks (as reaching a given point while avoiding obstacles). The algorithm can be tailored based on the communication and computational capabilities of the robots. A discussion and possible counterexamples of the applications of existing approaches are provided in the paper. Numerical simulations illustrate the results.