Assessment of Perceived Level of Service of Safety for Two-Lane Rural Highways: Case Study of Indian Bus Drivers

The choice of driving speed is a function of how safe a driver feels while driving in a particular road segment, and this is influenced by the road and roadside conditions. Researchers tried to study the effect of various road elements like road geometry, lane width and condition, shoulder width and condition and various roadside elements and hazards like presence of poles, trees, etc., on choice of driving speed for various class of vehicles. However, limited work attempted to associate various road alignments, shoulder conditions and parking conditions with the level of driving safety provided by the two-lane rural highways for bus drivers. This paper aimed to study the effect of road alignment, shoulder condition and parking on speed and perceived safety of two-lane rural highways using field and subjective driving experiments for Indian bus drivers. Effects of the road and shoulder conditions on speed and perceived safety ratings were analysed using multivariate analysis of variance (MANOVA). The work also aimed to classify highway segments into five level of service of safety (LOSS) groups based on driving speed and safety perception using centroid-based machine learning clustering algorithm, K-Means clustering algorithm, and associate different road alignment, shoulder conditions and parking conditions with different LOSS groups. The LOSS groups provide an insight into the degree to which road alignment, shoulder condition and parking condition influence the safety perceptions of highways. The LOSS groups also provide an insight into the actual driving speed range and safety rating provided by the driver for different conditions of a two-lane highway. The LOSS classification provided may be used to design customized strategies for safety improvement of two-lane highway stretches.