Studying Driving Behavior on Horizontal Curves using Naturalistic Driving Study Data

Abstract

Many studies have been conducted to develop models to predict speed and driver comfort thresholds on horizontal curves, and to evaluate design consistency. The approaches used to develop these models differ from one another in data collection, data processing, assumptions, and analysis. However, some issues might be associated with the data collection that can affect the reliability of collected data and developed models. In addition, analysis of speed behavior on the assumption that vehicles traverse horizontal curves at a constant speed is far from actual driving behavior. Using the Naturalistic Driving Study (NDS) database can help overcome problems associated with data collection. This paper aimed at using NDS data to investigate driving behavior on horizontal curves in terms of speed, longitudinal acceleration, and comfort threshold. The NDS data were valuable in providing clear insight on drivers’ behavior during daytime and favorable weather conditions. A methodology was developed to evaluate driver behavior and was coded in Matlab. Sensitivity analysis was performed to recommend values for the parameters that can affect the output. Analysis of the drivers’ speed behavior and comfort threshold highlighted several issues that describe how drivers traverse horizontal curves that need to be considered in horizontal curve design and consistency evaluation.

Modeling Operating Speed Using Continuous Speed Profiles on Two-Lane Rural Highways in India

“…Recent studies in the US evaluated driver behavior on horizontal curves using NDS (Naturalistic Driving study) data conducted through the Strategic Highway Research Program (SHRP 2), which forms the largest dataset up to date (Dhahir and Hassan 2018). Dhahir and Hassan (2019) studied driver behavior on horizontal curves and tangents using NDS.…”

Evaluation of Vehicle Speed with Impact of Vehicular Driver Behaviour at Horizontal Curves

“…The results concluded that reducing vehicular speeds based on the optical circles is an effective tool to improve the driving behaviour of attentive nature to reduce crashes at horizontal curves (Awan et al, 2019). Moreover, recent studies have explored driver behaviour at horizontal curves using the Naturalistic Driving Study (NDS) and the results concluded that the importance of speed behaviour as well as driver comfortable thresholds while traversing horizontal curves needs to be considered to maintain design consistency (Dhahir and Hassan, 2018). Whereas, in another NDS research, has shown that the driver's familiarity with the curves and the results identified that driver familiarity has a significant role in choosing higher vehicle speeds along the curves (Pratt et al, 2019).…”

Hierarchical quantitative analysis to evaluate unsafe driving behaviour from massive trajectory data

“…Toledo and Lotan used IVDR called DriveDiagnostics system to capture the correlations between driving risk indices and car crashes, and the results suggest that a series of driving risk indices could be employed to evaluate the driving safety and the risk of involvement in car crashes [19]. Dhahir and Hassan proposed a methodology to evaluate driving behaviour based on NDS database, and further to analyse the horizontal performance on driving experience, including speed, longitudinal acceleration, and comfort threshold [20]. As an important traffic data, trajectory data has drawn considerable attention in the intelligent transportation area [21], such as traffic congestion estimation, road network discovery, driving destination prediction etc.…”