- Núñez

In the context of MRTA, I studied how to attain an optimum balance between coverage and sensing. I obtained optimal solutions under various cost schemes [1,2,5] that rely only on local information. Unlike previous results in the area, I consider both tasks that are of effectively different nature, and different realizations of the same task simultaneously. By looking at coordination problems, I improved the existing algorithms for two of the most common distributed robotic tasks: coverage and rendezvous. For coverage, I extended the approach from Cortés et. al. [10] to non-convex regions [4,6]. This generalization is highly relevant since most real world environments are non-convex. My result thus makes coverage algorithms practical for real world scenarios, including homeland security, environmental monitoring and ocean pollution control. Moreover, since the coverage problem can be shown to be equivalent to the quantization problem solved by Lloyd’s algorithm in signal processing, my extensions now allow its implementation in applications which are not restricted to a convex space. As for rendezvous, I solved the general problem by exhibiting it as a particular realization of consensus algorithms. My solution [3,8] describes how the rich literature on consensus protocols can be applied to analyze and solve the rendezvous problem, and also shows how most of the previous approaches to rendezvous are a particular realization of my solution. Although the relationship between rendezvous and coverage was not unnoticed in the literature, it was not until my paper [3] that an explicit description of the nature of this relationship was presented.