Finding Frustration: a Dive into the EEG of Drivers

Kurzfassung

Emotion recognition technologies for driving are increasingly used to render automotive travel more pleasurable and, more importantly, safer. Since emotions such as frustration and anger can lead to an increase in traffic accidents, this thesis explored the utility of electroencephalogram (EEG) features to recognize the driver’s frustration level. It, therefore, sought to find a balance between the ecologically valid emotion induction of a driving simulator and the noise-sensitive but highly informative measure of the EEG. Participants’ brain activity was captured with the CGX quick-30 mobile EEG system. 19 participants completed four different frustration-inducing and two baseline driving scenarios in a 360° driving simulator. Subsequently, the participants continuously rated their frustration level based on the replay of each scenario. The resulting subjective measures were used to classify EEG time periods into episodes with or without frustration. Results showed that the frequently used measure of the Alpha Asymmetry Index (AAI) had, as hypothesized, significantly more negative indices for high frustration (vs. no frustration). However, a commingling effect of anger on this result could not be dismissed. The results could not provide evidence for the yet to be replicated previous research of frustration correlates within narrow-band oscillations (delta, theta, alpha, and beta) at specified electrode positions (frontal, central, and posterior). This thesis concludes with suggestions for subsequent research endeavors and forthcoming practical implications in the form of insights acquired.