Comprehensive Analysis of Distracted Driving Crashes among At-Fault Commercial Motor Vehicles in Kentucky Using Multiple Techniques

Abstract

There is a lack of studies that analyzed distraction-related crashes involving commercial motor vehicles (CMVs) (especially with CMVs being the at-fault vehicle). This study comprehensively investigates the contributory factors associated with severe (i.e., fatal and suspected serious injury) distraction-related CMV crashes (with CMVs as the at-fault vehicle) in Kentucky, 2019–2022. This study investigated crash, roadway, and rarely explored real time weather variables in CMV studies (including solar radiation, relative humidity, air temperature, wind speed, precipitation, and visibility) that were collected from the High-Resolution Rapid Refresh database within 1 h of the crash. Multiple statistical approaches were applied, including the association rules mining (ARM) (to uncover associations/interdependencies between the variables and distraction-related CMV crash severity), odds ratio (OR), and Chi-square (χ²) test of independence. The OR results revealed that distraction-related CMV crashes involving head-on collisions, speeding, curved roads, presence of vertical gradient, foggy weather, relative humidity (≤ 65%), and solar radiation (≥ 200 W/m²) had the highest severity odds. The ARM technique validated the preliminary analyses (using the χ² test and OR) and showed 21 interesting rules, of which, air temperature (≥ 70°F), no precipitation or rainfall, visibility (≥5 mi), presence of vertical gradient, and interaction between “presence of vertical gradient and relative humidity ≤ 65%” were significantly associated with increased serious injury likelihood. Based on this study’s findings, the installation of dynamic message signs when relative humidity is ≤ 65% in advance of vertical gradients and the installation of rumble strips will help reduce the severity of CMV distraction-related crashes.