The predominantly urban roads of the ACT create a complex environment in which drivers must quickly detect and respond to changing hazards. This project comprised three experiments designed to assess factors that affect drivers’ ability to detect changes in visual information and specifically exploring whether sleepiness impairs change detection, as no previous published research had examined this. Experiment 1 assessed factors that affect drivers’ change detection using photographic stimuli representing urban and rural driving scenes. Accuracy, response time (RT) and eye movements were measured. Participants showed superior change detection in rural compared with urban scenes, and for changes involving road users, animals and traffic lights, compared to inanimate objects (signs and trees). Experiment 2 used a modified version of the Experiment 1 task to explore the effect of sleep loss on change detection. Participants completed the change detection task twice, once after a normal night’s sleep (8 hours) and once following a night of sleep restriction (5 hours). Sleepiness did not impair accuracy, but was associated with increased RT to detect changes in urban scenes. As in Experiment 1, participants were more efficient at detecting changes to other road users than static objects (trees and signs) and were better at detecting changes in rural scenes compared to urban scenes. Experiment 3 was conducted in the CARRS-Q advanced driving simulator. Participants’ ability to detect expected and unexpected changes while driving in simulated urban and rural areas was compared when alert (8 hours sleep) and sleepy (5 hours sleep). Sleep loss did not significantly impair detection of expected changes; however, there was a non-significant reduction in detection of unexpected changes. Participants were bett r at detecting changes with high safety relevance and in urban areas (where travel speed was low), compared to rural areas (where travel speeds were high). Overall, this research suggests that drivers are better at detecting changes that involve other road users and targets with high safety relevance. The impact of safety relevance is greatest in demanding situations, e.g. when the visual environment is cluttered or at high travel speeds. There is limited evidence that sleep loss impairs efficiency of change detection in visually cluttered urban scenes. Future research is necessary to understand the vulnerability of visual attention to sleep loss.
Project Final Report Prepared for the NRMA-ACT Road Safety Trust / p.81