Assessment of Drowsy Driving Associated Characteristics Using Deep Learning

Abstract

The presence of road traffic accidents is subjected to various contributing factors including drowsy driving. The occurrence of drowsy driving has been a major cause of road accidents globally. Therefore, this study aims to analyze demographic, socio-economic, daily habits and drowsy-related characteristics associated with fatigued or drowsy driving in the United Arab Emirates (UAE). The data were gathered upon a questionnaire-based survey among a sample size of 525 drivers in the UAE. Inputs were given weights upon consulting experts in the field of transportation. Data were analyzed using artificial neural networks (ANN). Daily habits significantly affect the driver’s risk to experience fatigued driving. Socio-economic, drowsy-related, and demographic characteristics followed sequentially. Time of day to experience drowsy driving has the largest importance. Moreover, daily habits such as driving durations, distance driven, and sleeping hours demanded the importance of drowsy driving risk next. Socio-economic characteristic such as the average monthly income was the least significant. Prevalence of sleep-related accidents in the UAE is a fact, where drivers are less concerned about fatigue driving than other traffic safety issues. Raising awareness of drowsy driving among society is a need since people tend to see other factors to be riskier than drowsy driving. The results highlight the need to counteract drowsy driving with treatments on-road and more education to the public.