European Transport Research Review (May 2017)
The severity of driver fatigue in terms of line crossing: a pilot study comparing day- and night time driving in simulator
Abstract
Abstract Introduction The overall aim of this study is to compare day-time driving with night-time driving looking at line crossings during self-reported sleepiness and long blinks. The hypothesis is that high levels of self-reported sleepiness (KSS 9) and long blink duration (>0.15 s) will be less associated with critical events during the day-time compared to night-time. Method The study is based on data from a driving simulator experiment with 16 participants driving 150 km on a typical Swedish motorway scenario twice: once during daytime and once during night time. In total data from 6 segments of 4 km each equally distributed along the drive was averaged and included in the analysis. A Mixed Model Anova was used to test the effects on KSS, Blink Duration and Line Crossings with factors for Session (Day/Night) and Road segment (1–6), and participant as random. In addition, a logistic regression was used to identify when there is a risk for line crossings. Finally, the proportion of line crossings in relation to high KSS values and long blink durations was tested with Fisher’s exact test. Results The results show no differences in the percentage of Line Crossings to the left during high levels of Karolinska Sleepiness Scale during daytime (33%) compare to night-time (40%). However, there was a significant difference between day and night time line crossings while the driver had long duration blinks (4% during daytime and 35% during night-time). Despite these results the most promising predictor of line crossings in each segment of 4 km/h was KSS with an Odds Ratio of 5.4 with a reference value at Karolinska Sleepiness Scale level 5. Conclusion In conclusion, the results do not support the hypothesis that high levels of KSS will result in more frequent line crossings at night time compared to day time. However, the result supports the hypothesis that long blink durations are associated with more line crossings when they appear during night time than during daytime.
Keywords