@article{ref1,
title="Diagnosis of encroachment-related work-zone crashes by applying pattern recognition",
journal="Transportation research record",
year="2023",
author="Das, Subasish and Tabesh, Mahmood and Dadashova, Bahar and Dobrovolny, Chiara",
volume="2677",
number="7",
pages="222-236",
abstract="Work-zone safety is one of the critical goals of transportation agencies. Vehicles are required to change travel paths and lanes over a short length of a road section at work zones. Distracted drivers, unable to see advanced warning signals and pavement markings delineating the work-zone travel paths, could increase the likelihood of a crash. Recent statistics showed that fatal collisions in work zones had increased by 46% in 2019 compared with 2011. The frequency of roadway departures at work zones, the higher risk of fatalities, and little insight into encroachment types at work zones underscored the need for a thorough study. This study aimed to examine the vehicle encroachment conditions associated with work-zone locations and focused on 4 years (2016 to 2019) of crash data from the Texas Department of Transportation by applying a unique data-mining method known as cluster correspondence analysis. This method identified four clusters in both &quot;no-injury&quot; and &quot;fatal and injury&quot; crash data. Major factors contributing to vehicle encroachment were identified. Three dominating clusters were median-related crashes on two-lane divided high-volume roadways; single-vehicle overturning collisions on two-way divided roadways with unprotected median; and overturning crashes on two-lane undivided roadways in controlled traffic. The findings of this study will be useful for safety engineers to contribute to reducing encroachment-related work-zone crashes.<p /> <p>Language: en</p>",
language="en",
issn="0361-1981",
doi="10.1177/03611981231152254",
url="http://dx.doi.org/10.1177/03611981231152254"
}