Effects of Partial Sleep Deprivation on Braking Response of Drivers in Hazard Scenarios

Kirti Mahajan,Nagendra R Velaga
DOI: https://doi.org/10.1016/j.aap.2020.105545
Abstract:This study aimed at modeling the Response Time (RT) and Total Braking Time (TBT) of drivers under Partial Sleep Deprivation (PSD). Fifty male participants drove the driving simulator in three experimental conditions: two test sessions and a baseline. The two test sessions were conducted after one and two nights of PSD (sleep = 4.25 ± 0.5 h), respectively. Sleep reduction was recorded using a wrist-worn Actiwatch. The baseline session was conducted after full rest (7-8 h sleep/day for a week). The order of test sessions and baseline was randomized. Each test included two hazard events: 1) pedestrians crossing a road and 2) parked vehicles merging into a roadway. Karolinska Sleepiness Scale (KSS) and Sleepiness Symptoms Questionnaire (SSQ) ratings were also recorded during each drive. Four separate models using parametric accelerated failure time (AFT) with Weibull distribution were developed for RT and TBT in the two events. The models were chosen with clustered heterogeneity to account for intra-group heterogeneity due to repeated measures across tests. In the case of pedestrians crossing event, RT increased by 10% in the first test session and no significant effect observed on RT in the second test session. The overall TBT reduced by 25% and 28% during the first and second PSD sessions, respectively. In the case of vehicle merging event, both response time and total braking time delayed by 44% and 17% respectively after PSD. Other factors such as age, experience, work-rest hours, KSS and SSQ rating, often exercising, approaching speed and braking force were also found significant in the analysis. The parametric AFT approach adopted in this study showed the change in 'response time' and 'total braking time' concerning the type of hazard scenario and partial sleep-deprivation.
What problem does this paper attempt to address?