Severity of traffic accidents in Ecuador

Abstract

This research aims to analyze the severity of traffic accidents in Ecuador by identifying intervening factors, and applying statistical models in the identification of accident patterns. The study focused on 94 traffic accidents involving buses between 2017 and 2018. A methodology was used that included the analysis of variables grouped into four main categories: infrastructure, vehicle, environment and human factor. Through CART and Random Forest Decision Trees, the most significant variables were determined to analyze the severity of the accident. Among the results, it was determined that the severity associated with the type of natural shoulder is high with a probability of 92.3%, likewise, when the place of the accident is on a curve or straight, the severity associated with the type of accident such as the reach of the bus, overturning on the roadway, leaving the road and overturning is high with 83.3%. In conclusion, the study highlights that the accident patterns identified as highly severe traffic accidents in the Ecuadorian context include situations where the shoulder is in poor condition, the accident manifests itself as a rollover and occurs in curved sections of the road.