Detection of a driver's visual attention using the online UFOV method

This paper investigates whether a driver's useful field of view (UFOV) can be measured during normal driving tasks without the driver being aware of the system. We were able to develop an online method to measure the driver's UFOV (oUFOV) automatically using a real car driving test, in the city of Tampere. The system was merged with two different technical solutions that estimated if the driver paid attention to the traffic objects that appeared in his or her peripheral vision. Here, the original laboratory method using a PC display was expanded to monitor the driver for much longer periods. By incorporating the oUFOV method into a driving simulation environment in Japan, we were also able to conduct safer tests for the drivers with reduced visual conditions, such as the driver's mental status change (hasty driving), and replace more accurate control parameters with simulation software in cases of risky driving scenarios. Thus, we confirmed that the oUFOV can measure the visual conditions for the driver and be deployed as a real-time driver support system with an automatic intrusive HMI adaptation function or immediate alarming, in the future, without driver awareness, according to the driver's visual attention capability.