1: \begin{abstract}                % Abstract of not more than 250 words.

2: This paper regards coordination of networked systems, which is studied in the framework of hybrid dynamical systems.  We design a coordination scheme which combines the use of ternary controllers with a self-triggered communication policy. The communication policy requires  the agents to collect, at each sampling time, relative measurements of their neighbors' states: the collected information is then used to update the control and determine the following sampling time. We prove that the proposed scheme ensures finite-time convergence to a neighborhood of a consensus state. 

3: %We estimate both the size of this error neighborhood, and the time and communication costs required to reach it, as functions of a sensitivity parameter: our analysis highlights natural trade-offs between costs and accuracy in approximating consensus. 

4: %

5: We then study the robustness of the proposed self-triggered coordination system with respect to skews in the agents' local clocks, to delays, and to limited precision in communication.

6: %

7: Furthermore, we present two significant variations of our scheme. First, we design a time-varying controller which asymptotically drives the system to consensus.

8: Second, we adapt our framework to a communication model in which an agent  does not poll all its neighbors simultaneously, but single neighbors instead. 

9: This communication policy actually leads to a self-triggered ``gossip'' coordination system.

10: \end{abstract}

11: