Systematic Evaluation of Driver’s Behavior: A Multimodal Biometric Study

Divya Seernani

Michela Minen

Luisina Gregoret

Jessica Wilson


Complex traffic areas and high cognitive workload while driving are leading contributors to traffic crashes. Even though cognitive workload and stress have been previously assessed through various neurophysiological responses, they are rarely characterized simultaneously, limiting the triangulation of behavioral metrics (like drivers’ visual attention and facial coding) with physiological measures to investigate their interplay. The aim of the present study was to systematically characterize stress and cognitive workload through a multimodal assessment comprising eye-tracking, facial expressions, galvanic skin response (GSR), electromyography (EMG), electrocardiography (ECG) and respiration in three controlled driving simulations of varying complexity: 1. Baseline driving on an open road (Baseline); 2. Navigating between traffic cones (Cones); and 3. Driving in a neighborhood with multiple stressors (Traffic). The selected metrics were eye tracking dwell time, the presence of facial brow furrow, GSR peaks/minute, EMG activity of the upper trapezius muscle, heart rate and heart rate variability (HRV) and respiration cycles/minute. Physiological responses showed significant increases in GSR, heart rate and trapezius EMG activity with Cones and Traffic compared to Baseline. Eye tracking metrics were shown to be indicative of driving behavior in different conditions. There were no significant differences in facial expressions, HRV or respiration. These results are somewhat consistent with previous literature, suggesting that a multimodal approach to physiological signals can characterize affective and cognitive states in driving scenarios.


  • Driving simulator
  • Cognitive workload
  • Eye tracking
  • Physiology
This publication uses ECG, EMG, Eye Tracking Glasses, Facial Expression Analysis, GSR and Respiration which is fully integrated into iMotions Lab

Learn more