Acclimatizing to automation: Driver workload and stress during partially automated car following in real traffic

Abstract

Automated driving systems are increasingly prevalent on public roads, but there is currently little knowledge on the level of workload and stress of drivers operating an automated vehicle in a real environment. The present study aimed to measure driver workload and stress during partially automated driving in real traffic. We recorded heart rate, heart rate variability, respiratory rate, and subjective responses of nine test drivers in the Tesla Model S with Autopilot. The participants, who were experienced with driver assistance systems but naïve to the Tesla, drove a 32 min motorway route back and forth while following a lead car in regular traffic. In one of the two drives, participants performed a heads-up detection task of bridges they went underneath. Averaged across the two drives, the participants' mean self-reported overall workload score on the NASA Task Load Index was 19%. Moreover, the participants showed a reduction in heart rate and self-reported workload over time, suggesting that the participants became accustomed to the experiment and technology. The mean hit (i.e., pressing the button near a bridge) rate in the detection task was 88%. In conclusion, driving with the Tesla Autopilot on a motorway involved a low level of workload that decreased with time on task.

Effects of platooning on signal-detection performance, workload, and stress: A driving simulator study

“…Standardized change scores for each scale of the DSSQ were calculated as follows: (post-score À pre-score)/(standard deviation of the pre-score) (Helton et al, 2002). The scores for the three DSSQ scales (engagement, distress and worry) were calculated as the means of four subscales (based on Fairclough and Venables, 2005;Heikoop et al, 2016b;Matthews et al, 2002;Matthews, 2014). Specifically, engagement consists of the subscales (1) energetic arousal,…”

Investigating the Significant Individual Historical Factors of Driving Risk Using Hierarchical Clustering Analysis and Quasi-Poisson Regression Model

“…The study inferred that speed maintenance and lane deviations along with driving experience and gender groups may play an important role in analyzing and improving traffic safety in highway environments. Some research has analyzed the driver's workload of physiological signal [9,10]. In [9], the authors investigated the psychological effects of stress and workload on drivers' behavior during automated driving.…”

“…Some research has analyzed the driver's workload of physiological signal [9,10]. In [9], the authors investigated the psychological effects of stress and workload on drivers' behavior during automated driving. The study [10] proposed a method for assessing the driver's workload in real context and for obtaining the objective index of the drivers' experienced workload to provide a potential reliable online trigger.…”

A Novel Classification Method for a Driver’s Cognitive Stress Level by Transferring Interbeat Intervals of the ECG Signal to Pictures

“…This type of method is reliable because the physiological data can directly reflect a driver's physiological state and will change when the driver's attention changes, and the process of signal acquisition is more convenient and easier with the development of wireless signal acquisition technology. Researchers in this field are paying more attention to the application of physiological signals, such as electrocardiogram (ECG) [13][14][15], electroencephalogram (EEG) [16][17][18][19][20][21], galvanic skin response (GSR) [22][23][24][25], and electrooculogram (EOG) [26][27][28]. Improving the accuracy and reliability of recognition is a research goal of scholars.…”