Abstract
Purpose :
The visual challenges encountered in night driving involve low light levels, oncoming headlights and adaptation to fluctuating light levels. It is important to characterize these factors, given the complex and evolving nature of the modern night-time driving environment. This study examined the variations in pupil size and light levels encountered in typical urban, night-time roads.
Methods :
Nine participants (aged 31.1 ± 9.6 years) drove a 16-km in-traffic urban route at night, which included several road types (expressway, high-traffic arterial and residential roads). Pupil size was continuously measured during the drive with an eye-tracker using its in-built infrared pupil detection algorithm (Pupil Labs, 200Hz sampling rate), while an illuminance sensor (EKO Instruments, 2Hz sampling rate) attached to the eye tracker measured the vertical illuminance at the eye. Linear mixed models explored the variations in at-eye illuminance and pupil size while driving in the road types. Pupil size was also measured under laboratory mesopic (1cd/m2) and photopic (140cd/m2) conditions (NeurOptics PLR-200).
Results :
The mean ± 1SD at-eye illuminance on residential roads was 0.5 ± 0.7 lux, and was significantly lower than expressway (2.5 ± 1.5 lux) and arterial roads (3.2 ± 2.3 lux) (p<.001). Mean pupil size while driving on residential roads was 5.6 ± 0.5 mm, significantly larger than on the expressway (5.2 ± 0.4 mm) and arterial roads (4.9 ± 0.5 mm) (p<.001). Variability in at-eye illumination, based on the SD, was lowest on residential roads compared to the other roads (p<.001), yet this pattern was not reflected in pupil size variability, which was similar between residential and arterial roads, and lowest on expressways (p<.01). Pupil size when driving was typically closer to laboratory mesopic pupil size than photopic, but considerable fluctuations in these values was observed, both between different road categories and participants.
Conclusions :
This study provides information on levels of at-eye illuminance experienced by drivers in the modern night-time driving environments, and associated pupil responses. There is considerable variability in pupil responses and this is not reflected in typical laboratory-based measures of pupil size. This approach provides a novel strategy for exploring night-time driving difficulties with optical devices, and informs vision testing protocols to reflect the visual challenges of night-time roads.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.