Connectivity of Collaborative Robots in Partially Observable Domains

Abstract

Abstract: Collaborative information processing is vital for a swarm of robots tasked with many different applications. Swarm connectivity is necessary for achieving good collaboration. Our approach to this is a physics-based autonomous robot framework that acts as a distributed mobile sensor network that is capable of maintaining high connectivity during self-organization and movement. The framework, called Physicomimetics, is a robust control scheme built on local interactions between the robots, making it highly scalable, adaptive, and cost effective. This paper presents connectivity results of mobile sensor networks on two partially observable domains – formation movement through obstacle fields and the self-organization of chain formations.