AUTHOR=Giorgi Andrea , Ronca Vincenzo , Vozzi Alessia , Aricò Pietro , Borghini Gianluca , Capotorto Rossella , Tamborra Luca , Simonetti Ilaria , Sportiello Simone , Petrelli Marco , Polidori Carlo , Varga Rodrigo , van Gasteren Marteyn , Barua Arnab , Ahmed Mobyen Uddin , Babiloni Fabio , Di Flumeri Gianluca 

TITLE=Neurophysiological mental fatigue assessment for developing user-centered Artificial Intelligence as a solution for autonomous driving

JOURNAL=Frontiers in Neurorobotics

VOLUME=Volume 17 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/neurorobotics/articles/10.3389/fnbot.2023.1240933

DOI=10.3389/fnbot.2023.1240933

ISSN=1662-5218

ABSTRACT=Human factor plays a key role in the automotive field, since most accidents are due to driver’s unsafe behaviors. To deal with this, two main solutions are pursued: on a short term the development of systems aimed at monitoring driver’s psychophysical state, such as inattention and fatigue; and on a medium-long term the full autonomous driving. This second objective is promoted by the recent technological progress in terms of Artificial Intelligence and sensing systems. The assumption is that the more accurate the vehicle will be aware of the “surroundings”, the more reliable will be the autonomous driving. 
Even with autonomous vehicles, the drivers should be able to take control of the vehicle when needed (i.e., takeover request), especially during this transition phase from lower (SAE < 3) to the highest level (SAE = 5) of autonomous driving.In this scenario, the vehicle has to be aware not only of the “surroundings” but also of the driver’s psychophysical state, i.e. a user-centered Artificial Intelligence. 
Neurophysiological approach is one the most effective in detecting users' unproper mental states while driving. This is particularly true if considering that the more automatic the driving will be, the less available the vehicular data related to the driver’s driving style will be. Several signals such as those related to brain, ocular and cardiac activities, have been largely adopted in scientific research to characterize episodes of mental fatigue or inattention. 
The present study aimed at employing a holistic approach, considering simultaneously electroencephalographic, electrooculographic, photopletismographic and electrodermal activity data, on 26 professional drivers engaged in a long-lasting realistic driving task in simulated conditions. The aim was to investigate which neurophysiological parameters can be used together to assess the driver’s mental fatigue in real time and to detect the onset of fatigue, to make this information available for the vehicle AI. Results showed that the most sensitive parameters are those related to brain activity. In a less extent, also those related to ocular parameters are sensitive to the onset of mental fatigue, but with a delayed effect. The other investigated parameters did not significantly change during the experimental session.