Spatial Analysis of Freeway Incidents and Incident Detection

This paper presents an analysis of the spatial distribution of incidents and incident detection on a traffic management center’s instrumented network. The researchers evaluated the spatial distribution of incidents according to different categories including (a) temporal distribution of incidents by: month and season (academic year versus summer), day of week and weekday versus weekend day, and time of day (AM peak, midday, PM peak, and evening and early-morning hours); (b) traffic volume; and (c) corridors. To conduct this analysis, the researchers used archived ITS data, particularly freeway incident and incident alarm data, as well as a geodatabase of ITS features. In addition, the researchers used several GIS tools to perform the spatial queries needed for the analysis. A number of maps were developed to provide sector-by-sector performance measures such as incident rates and incident detection rates. This analysis helps optimize freeway incident management practices. For example, it can be used to identify freeway segments with higher-than-normal incident rates and to accordingly optimize service patrol schedules and routes. Although the analysis described in this paper used data from one jurisdiction (San Antonio, Texas), the methodology is sufficiently generic to enable implementation at other traffic management centers.