ABSTRACT
Animal behavior is greatly influenced by interaction between peers as well as with the environment. Understanding the flow of information between individuals can help decipher their behavior. This applies to both the microscopic and macroscopic levels, from cellular communication to coordinated actions by humans. The aim of this work is to provide a simple but sufficient model of information propagation to learn from natural coordinated behavior, and apply this knowledge to engineered systems. We develop a probabilistic model to infer the information propagation in a network of communicating agents with different degrees of interaction affinity. Another focus of the work is estimating the time needed to reach an agreement between all agents. We experiment using swarms of robots to emulate the communication of biological and social media groups for which we are able to provide upper bounds for the time needed to reach a global consensus, as well as to identify individuals that are responsible for slow convergence.
