Assessing Driver Frustration using functional near infrared spectroscopy (fNIRS)

Abstract

Introduction and Aim: Driving is a goal-directed task. During traffic, blocking obstacles occur often eliciting driver frustration, and may result in more risky driving including speeding and aggressive behavior towards other traffic road participants. We envision to use ‘adaptive automation’ [1] to design driver assistance systems that prevent those maladaptive driving behavior by supporting the driver to reduce the current frustration level or increasing the level of automation. For this, it is essential that we can assess driver frustration already at an early stage. As affective states come along with a change in cognitive appraisal and a subjectively experienced feeling, brain activity seems to be a promising indicator of frustration. Thus, we aim to investigate if brain activation fNIRS measurements could be used to assess realistically occurring driver frustration. Methods: Thirteen volunteers participated in a driving simulator study, in which frustration was induced through a combination of time pressure and blocking events. A cover story asked the participants to imagine that they had to deliver a parcel to a client within six minutes with the incentive of gaining 2 € per successful delivery. In the six frustrating scenarios (FRUST), the drives were blocked by heavy traffic, while in the six non-frustrating scenarios (NOFRUST), participants could drive mostly unblocked. FNIRS data were recorded from the frontal and parietal cortices using a 32-channel neuroNIRX-system during the course of the whole experiment. Results: We subjected the individual fNIRS data to an SPM-based GLM analysis which revealed relative concentration increases of oxy-hemoglobin (HbO) and decreases of deoxy-hemoglobin (HbR) during FRUST compared to NOFRUST drives. These changes were found bilaterally in ventrolateral frontal and temporo-occipital channels in the group analysis as well as at the individual level (see Fig. 1 for group level and an example of an individual analysis. Note the much higher t-values in the individual analysis). Discussion & Conclusion: Our results indicate that brain activity patterns of frustrated drivers are discernible from non-frustrated drivers using fNIRS. The revealed brain areas displaying higher activity in the frustrating drives are in line with the relatively scarce literature on frustration-related neuroimaging results and seem to be related to cognitive appraisal, impulse control and emotion regulation processes. Since frustration is a multi-component phenomenon, it could be helpful to measure behavioral and other physiological measures in combination with fNIRS data to acquire a robust assessment of frustration. Acknowledgements: This work was funded by a grant of the Volkswagen Foundation and the Ministry of Science and Culture of Lower Saxony to the Research Centre on Critical Systems Engineering for Socio-Technical Systems. Klas Ihme and Anirudh Unni contributed equally. References [1] Parasuraman, R. (1987). Human-computer monitoring. Human Factors, 29, 695-706