Driving simulator studies have shown that when automated driving at SAE Level 2 is engaged, drivers spend more time looking away from the forward roadway, and less at safety critical locations, such as the rear-view and side mirrors, when compared to manual driving. A similar pattern of driver attention allocation is now observed on the real road, with L2 automated vehicles. This reduced engagement by the human in the driving task is known to take them “out of the control loop”, leading to a slower resumption of control of the driving task after a takeover request. Response to critical events is also shown to be slower, when compared to manual driving. As the level of automation in vehicles increases, and the driver is allowed to engage in other (non-driving related) activities, the challenge of keeping them suitably “ready” to take back control from the vehicle is even higher. This talk will provide an overview of a number of studies conducted at the University of Leeds, which have used camera-based driver monitoring systems (DMS) to understand how drivers’ visual attention is distributed during different stages of L2 and L3 automated driving, and how this affects resumption of control after a takeover request. The use of such DMS for confirming readiness, before the driver is allowed to take control, is also outlined. Results are discussed in relation to the use of different HMI for taking drivers’ attention back to the road in an efficient manner, and what must be considered for future developments of in-vehicles sensors to improve driver response and safety during higher levels of automated driving.