Detecting Drivers’ Stress in Emergency Using Physiological Sensors

Abstract

This paper presented methods for collecting and analyzing heart rate data during virtual-world driving task in emergency to distinguish and assess a driver’s stress state. The primer study had developed the driver stress training system loaded on driving simulator to display 14 typical emergency scenarios in different road environment. Electrocardiogram was recorded continuously by Physiological Sensors while drivers followed a random simulated driving on the driving simulator. Data from 30 drivers of at least 10 times per scenario were collected for analysis. The data were analyzed in two ways of HRV (Heart Rate Variability) analysis. Analysis I used time domain indexes of data during the driving in the whole training to distinguish the stress occur during the time frame of the scenario with an accuracy of over 98% across multiple drivers and driving days. Analysis II compared frequency indexes of data in stress state in emergency and in clam, with a metric of observable stressors created by the scenarios. The results showed that heart rate metrics is most closely correlated with driver stress level for most drivers studied. These findings confirmed that heart rate signals can provide a metric of driver stress in future cars capable of physiological monitoring. Such measurement could be used to help manage non-critical and critical in-vehicle information systems and could also provide a continuous measure of how different road and traffic conditions affect drivers. Physiological sensors can be used in driving assistance system in future.