Factors Associated with Stopping to Rest During Long Night-Time Simulated Drives

Method

Participants (N=90, M = 31.4 yrs, SD = 10.3) were recruited and completed the simulated nighttime drive in a miniSim fixed-base driving simulator. The simulated drive consisted of 150 miles of driving on an Interstate highway in dark nighttime conditions. The simulated drive included 7 identical rest areas spaced approximately 20 miles apart. Participants were told that they had the option to stop at rest areas during the drive but were not told of their locations or spacing. Upon exiting the simulator, participants could take a nap in a private room with a comfortable chair, use the restroom, walk around, step outside, and/or consume coffee (8 oz) or another caffeinated beverage. A monetary incentive was employed to attempt to replicate the motivational tradeoffs associated with driving while drowsy, i.e., to reach one’s destination safely but also quickly.

The main dependent variable was stopping at a simulated rest stop during the drive. Activities that participants performed during breaks were also recorded. A break was considered to have been motivated at least in part by drowsiness (hereafter drowsy break) if the participant self-reported drowsiness, sleepiness, fatigue, or similar as a reason for their break, or if they were observed napping or obtaining a caffeinated beverage. Independent variables included subjective and objective drowsiness and measures of time and driving performance. Subjective drowsiness was assessed using the Karolinska Sleepiness Scale (KSS) (Åkerstedt & Gillberg, 1990). Objective drowsiness was assessed using a modified PERCLOS that quantified the percentage of time that participants’ eyelids were fully closed, coded by trained research assistants. Other independent variables included standard deviation of lane position, total elapsed time since the beginning of the drive, time driving since last break, number of breaks, and rest area number.

Results

51% of drivers took at least one break. The timing of the first break ranged from 22-135 minutes into the drive. On average, the first break occurred a little over an hour into the drive (77min), although there was a cluster of breaks that occurred about 100 minutes into the drive. For drivers who took a second break, on average that break occurred 44 minutes after the first break and approximately 100 minutes into the total drive. A logistic regression model predicting drowsy break taking with the independent measures was statistically significant. KSS score was the only predictor variable significantly associated with an increase in the likelihood of taking a drowsy break. No other predictor variables, including both PERCLOS and SDLP measures, had a significant association with the likelihood of taking a drowsy break.

Discussion

Drowsy breaks were predicted by drivers’ self-ratings of drowsiness, measured by the KSS. KSS ratings were the only dependent measure that was a significant predictor of break-taking behavior. This finding suggests that, at least in the context of the incentive structure in this study, drowsy drivers rely on self-perceptions of their drowsiness when deciding whether to take a break or not. This is in line with the finding that KSS ratings were somewhat poorly calibrated with objective PERCLOS measurements (Gaspar & Carney, 2023). It is important to note that most drivers chose to continue driving despite rating themselves as drowsy. Even at the highest KSS rating (9), more than 75% of the time participants decided to continue driving rather than taking a break. Results underscore the need to educate and/or motivate drivers to respond sooner to warning signs of drowsiness.