Using naturalistic driving data to quantify driver following behavior during braking

Abstract

This study aims at quantifying and comparing drivers' following behavior during braking at intersections and midblock road segments (i.e., non-intersection) using Naturalistic Driving Data (NDS) data. The quantification of driver behavior was based on exploring the Probability Density Functions (PDF) of two parameters (i.e., minimum following distance and time-to-collision) which were extracted from NDS data. To classify drivers' following behavior at intersection and segments, all events were mapped onto the road network using ArcGIS. The intersection events were defined based on the driver's stopping sight distance to the stop line of an intersection. The results showed that there was a significant statistical difference between the PDFs of the parameters at intersection and segment events. Also, the results showed that drivers tend to keep less following distances and high deceleration rates (i.e., more aggressive) during intersection events when compared to segment events. Generally, the results of this study highlight the importance of considering the driver location before judging or classifying his/her behavior.