This paper addresses the classification problem for a set of autonomous robots that interact with each other. The objective is to classify agents that "behave" in "different way", due to their own physical dynamics or to the interaction protocol they are obeying to, as belonging to different "species". This paper describes a technique that allows a decentralized classification system to be built in a systematic way, once the hybrid models describing the behavior of the different species are given. This technique is based on a decentralized identification mechanism, by which every agent classifies its neighbors using only local information. By endowing every agent with such a local classifier, the overall system is enhanced with the ability to run behaviors involving individuals of the same species as well as of different ones. The mechanism can also be used to measure the level of cooperativeness of neighbors and to discover possible intruders among them. General applicability of the proposed solution is shown through examples of multiagent systems from Biology and from Robotics.

In this paper, we address decision making problems, depending on a set of input events, with networks of dynamic agents that have partial visibility of such events. Previous work by the authors proposed so-called logical consensus approach, by which a network of agents, that can exchange binary values representing their local estimates of the events, is able to reach a unique and consistent decision. The approach therein proposed is based on the construction of an iterative map, whose computation is centralized and guaranteed under suitable conditions on the input visibility and graph connectivity. Under the same conditions, we extend the approach in this work by allowing the construction of a logical linear consensus system that is globally stable in a fully distributed way. The effectiveness of the proposed method is showed through the real implementation of a wireless sensor network as a framework for the surveillance of an urban area.

Vision-based control of a robot formation is challenging because the on-board sensor (camera) only provides the view-angle to the other moving robots, but not the distance that must be estimated. In order to guarantee a consistent estimate of the distance by knowing the control inputs and the sensor outputs in a given interval, the nonlinear multi-robot system must preserve its observability. Recent theoretical studies on leader-follower robot formation exploit the interesting influence that the control actions have on observability. Based on these results, in this paper we present a switching active control strategy for formation control. Our control strategy is active in the sense that, while asymptotically achieving the formation control tasks, it also guarantees the system observability in those cases in which all the robots tend to move along non-observable paths. As a result, both estimation and formation performances are improved. Extensive simulation results show the effectiveness of the proposed design.