Understanding Drivers’ Emotion Variations During NDRTs in Conditional Automated Driving Using EEG Data

Drivers’ emotion is one of the key factors affecting driving behavior, negative emotion would be harmful to safe driving. In conditional automated driving, drivers would devote the majority of the time to Non-Driving Related Tasks (NDRTs), and varying NDRTs may lead to variations in drivers’ emotions. In this study, a series of conditional automated driving simulation experiments considering four typical NDRTs was conducted, and the electroencephalogram (EEG) signals were detected to identify and characterize drivers’ emotions by the valence-arousal (V-A) model, where in valence ranges from displeasure to pleasure and arousal ranges from relaxation to excitement. The results show that 1) the drivers’ emotion, as measured by the valence and arousal, was significantly affected by the rest and reading tasks during conditional automated driving, but not influenced by the monitoring vehicle operation and watching video tasks; 2) the V-A model revealed that the drivers were prone to keep alert in automated driving while showing increased stress in the rest state. The findings of this study contribute to a better understanding of drivers’ emotions and NDRTs in conditional automated driving.