Caracterización de los comportamientos de los usuarios viales por medio del empleo de VANT en intersecciones con mayor accidentalidad en Bogotá D.C.

Abstract

Road accidents has worldwide positioned as one of the leading causes of mortality, and Bogotá D.C., has not been the exception in the occurrence of this type of events. This document presents a geostatistical analysis using road accident information provided by the District Department of Mobility of Bogotá between years 2012 and 2015. The use of different geostatistical methods allowed an exploratory analysis of spatial data, as well as identification and selection of areas with high concentration of casualties, patterns of data distribution, trends and prediction of surfaces. This investigation estimates Hotspots, exploratory spatial data analysis AEDE and deterministic analysis with distance method IDW. These data were classified into five classes with next limits: Very low, low and very high. For the AEDE, and with Clustering and Moran index (high and low), it was possible to find the correlation between points without fatal victims and points with lower risk. The IDW index estimated an accident model. The combination of these elements provided a strong basis to classify the city into zones according to the accident rate of the city. These classification results will provide valuable spatial information for further research in order to focus measures against road accidents in specific areas with higher rates and to aid local authorities to take actions in those high risk areas. Additionally, field information was collected for the points determined by geostatistical methods, by means of videos using the UAV (unmanned aerial vehicles) focused on the capture and characterization of the most recurrent behaviors by the different transit actors. In addition, information was collected in the field on road infrastructure (vehicular and pedestrian), signaling, land use and traffic light phases, through this information were characterized the behaviors that are generated by the infrastructure of the site, which are associated with land uses and those that are subject to the behaviors typical of the drivers of the different vehicles as well as the pedestrians that pass through each sector. Once the information collected is analyzed, it is concluded that these geostatistical methods yield reliable results in terms of the location of concentration zones and the probability of occurrence of accidents, since through the characterization of behaviors a large number of highly dangerous events were determined for the generation of accidents.