Cooperative anchoring in heterogeneous multi-robot systems

Abstract

Highly heterogeneous robotic systems are becoming increasingly common, as are robotic systems integrated with smart environments. In such distributed systems, there are many different sources and types of information, which need to be coordinated and combined effectively. The problem of cooperative anchoring is (roughly) the problem of, in a distributed system, determining which items of information refer to the same objects, and combining these items accordingly. In this paper, we define a general computational framework for cooperative anchoring inspired by work on conceptual spaces and (single-robot) perceptual anchoring. We also discuss an implementation of this framework which uses tools from fuzzy logic, and we present an illustrative experiment.