Multi-agents adaptive estimation and coverage control using Gaussian regression

The continuous progress on hardware and software is allowing the appearance of compact and relatively inexpensive autonomous vehicles embedded with multiple sensors (inertial systems, cameras, radars, environmental monitoring sensors), high-bandwidth wireless communication and powerful computational resources. While previously limited to military applications, nowadays the use of cooperating vehicles for autonomous monitoring and large environment, even for civilian applications, is becoming a reality. Although robotics research has obtained tremendous achievements with single vehicles, the trend of adopting multiple vehicles that cooperate to achieve a common goal is still very challenging and open problem. In particular, an area that has attracted considerable attention for its practical relevance is the problem of environmental partitioning problem and coverage control whose objective is to partition an area of interest into subregions each monitored by a different robot trying to optimize some global cost function that measures the quality of service provided by the monitoring robots. The "centering and partitioning" algorithm originally proposed by Lloyd [1] and elegantly reviewed in the survey [2] is a classic approach to environmental partitioning problems and coverage control problems.