Dynamic Coverage via Multi-Robot Cooperation

Abstract

The traditional coverage problem (Gage, 1992), in the context of multi-robot teams, has been defined as the maximization of the total area covered by the sensors of all robots in the team. General coverage algorithms (O'Rourke, 1987; Howard et al., 2002; Batalin and Sukhatme, 2002) typically apply certain dispersion techniques (e.g. Planning, Potential Fields, Molecular Dispersion) to a group of robots. The system usually converges to a static configuration with a high degree of sensor coverage. In some cases the arrangement of robots may be such that the coverage is maximized. Single-robot exploration algorithms (Yamauchi, 1997; Zelinsky, 1992; Dudek et al., 1991) while explicitly addressing coverage, are closely related in that their aim is for one robot to completely discover its environment.