Neurometric Evaluation of Driver Cognitive Workload Across Different Simulated Highway Scenarios

Neurometric Evaluation of Driver Cognitive Workload Across Different Simulated Highway Scenarios Modern Advanced Driver Assistance Systems rely on reactive kinematic signals rather than the biological states that precede them, often flagging danger only after a driver is already degraded. This thesis introduces a synchronized biometric architecture built to collect GSR and EEG data from Shimmer and Mindtooth devices in real time while participants perform a simulated driving task, phase-locking both streams onto a shared timeline alongside eye-tracking data. Work-zone driving drove a marked rise in parietal Theta power, while steering corrections and skin-conductance spikes reliably preceded peak EEG stress evidence of a reactive, subcortical arousal response. A derived Sleepiness Index showed drowsiness rising once drivers moved from heavy traffic into free-flow conditions. A Feature Importance Analysis revealed an "Index Paradox": road geometry dominates failure prediction in work zones, while in free-flow cruising the driver's neurological state becomes dominant instead. This thesis delivers a validated multimodal framework for predictive, biology-informed ADAS intervention.