Identification of Factors Related to Violation Propensity Using Large Naturalistic Intersection Approach-Level Database

Zachary R. Doerzaph, Rajaram Bhagavathula, Feng GuoTransportation 2010

Abstract

It is estimated that each year 260,000 red light running crashes occur in the US, resulting in 750 fatalities. Furthermore, thirty percent of intersection related fatalities are associated with signalized intersections; this is high considering that only 10% of the nation's intersections are signalized. The primary objective of this paper is to identify risk factors that are related to the likelihood of red-light violations; with the ultimate goal of development of strategies to reduce intersection crashes. Potential violation risk factors were investigated using a case-control study design. Data collected with a highly-capable advanced data collection system provided high resolution data on potential risk factors included vehicle type, location, adjacent vehicle crossing types, weather, lighting conditions, the difference initial speed and the posted speed limit, traffic volume, and the time to intersection when the yellow phase was first presented. The logistic regression model was applied to a large database including nearly 6,000 violating and compliant intersection approaches. The significant factors identified included weather condition (if clouds are present a violation is 6 times more likely than if the sky is clear) and vehicle type (during a straight crossing path maneuver a heavy vehicle is over 3 times more likely to violate than a light vehicle).
BibTeX
@inproceedings{doerzaph2010identification,
  author = {Zachary R. Doerzaph and Rajaram Bhagavathula and Feng Guo},
  title = {{Identification of Factors Related to Violation Propensity Using Large Naturalistic Intersection Approach-Level Database}},
  booktitle = {Transportation},
  year = {2010},
}

Loading related content...

Figure