Persistent Awareness-Based Multi-Robot Coverage Control

Abstract

The persistent coverage control problem is solved in a decentralized fashion for a multi-robot team with limited sensing capability and loss of awareness over time. Each agent in the multi-robot team possesses a measure of awareness about the environment, encoded as time-varying density functions. The team is tasked with increasing their awareness of the domain in a decentralized way while coordinating with neigh-boring agents to maintain coverage over the entire domain. Local exchanges of awareness levels between neighbors is shown to allow awareness information over the entire domain to propagate to all agents with performance guarantees. A control law is proposed to obtain (locally) optimal coverage, and a bound of number of robots for guaranteed loss of awareness of the domain is provided. The proposed control law is verified in silico for a team of nonholonomic planar robots.