Antonella Ferrara Università di Pavia Traffic congestion is a critical problem in all the densely populated areas of the world. It significantly affects the quality of life of millions of citizens, involving high social/environmental costs. In general, the problem cannot be solved by building new infrastructures because of the existence of hard constraints, so that the extensive exploitation of advanced technology and efficient control systems is often the unique attractive/viable solution. Control strategies for freeway traffic systems have been studied for decades. Freeways can be controlled in different ways: the most widely used control approach is ramp metering, according to which the flows of vehicles accessing the freeway from the on-ramps are regulated by means of traffic lights. Ramp metering is usually applied to achieve optimal freeway traffic conditions. The traffic control approach normally used to pursue optimality is Model Predictive Control (MPC). Unfortunately, three main issues affect the practical applicability of MPC schemes to freeway traffic control: 1. a non-linear optimal control problem for a large scale system must be faced; 2. this highly complex problem must be solved at any control sampling step; 3. with the same rate, the system state measurements and the control signals have to be transmitted through the communication network connecting the sensors to the control center, and this latter to the actuators. Depending on the available technological infrastructure, the foregoing three issues can have a different impact on the actual applicability of the control scheme, but, typically, it is observed that the real-time application of MPC schemes is feasible only for small traffic networks. Several alternative solutions to overcome the drawback have been investigated in recent years. This talk will be focused on one of them, which seems particularly promising: the event triggered-control of freeway traffic. The basic idea is that of relying on suitable triggering conditions, based on the traffic system state, to decide when the state measurements actually need to be transmitted to the controller, thus enabling a re-computation, i.e. an update, of the control action. The talk will also describe model-based event-triggered traffic control schemes, in which the structure of a model-based networked control system is enriched with the use of event-triggered conditions in the sensors and in the controller. Simulation examples will be discussed.