Neural response based behavioral profiling of vehicle drivers to personalize alarm sequences, warn safety systems and trigger non-driver-in-loop control

Abstract

Quantitative measurement of vehicle driver alertness and response readiness to on-road emergency cues could add more response time to driver assistance and safety features. Exogenous Electroencephalogram (EEG) potentials generated in Beta frequencies can imply driver’s attention to the situation, the absence of which may be associated with distraction or drowsiness. Detection of specific Alpha potentials during action-soliciting events may further indicate risk of violating response-trigger thresholds beyond which physical and vehicular response periods may become too short to avoid collision. Specific Event-Related Potentials (ERP) associated with early automatic cognitive process of consciousness to obstacles, emotional reaction to visual cues, risk evaluation, drier’s emotional intensity, anticipation and motor preparation can be identified, isolated and processed to create a model that predicts driver’s intension/capability to respond to obstacles within a derived time threshold. Isolating signals that indicate driver in-activity (extreme fatigue, sleep, visual or auditory distraction) during near-obstacle situations could further be used to preempt user-in-loop drive, in semi-autonomous vehicles, even if the driver is inhibiting normal transition to self-drive mode. Existing safety components, like airbags, could be issued pre-warnings about possible crash situations (currently they are actuated only upon impact), giving them critical additional time to get pre-actuation sequences triggered.