Crash Rate as a Function of Access Point Density and Traffic Flow

High crash rates on U.S. highways results in a considerable loss of human and economic resources. It is estimated that the U.S. loses about $10 billion every year in motor vehicle crashes in terms of fatalities, injuries, and property damage. Demand for access points from highways to properties has steadily increased due to rapid growth in local economies creating thereby compromises between accessibility and mobility or capacity and safety. On the other hand, the availability of access is essential to residential and commercial developments and often occurs at the cost of safety and traffic operations. This often directs the need to negotiate a compromise between accessibility, mobility and safety. This research establishes a methodology that examines the best possible spacing between access points and geometric roadway factors that improves traffic flow and reduces traffic crash rates. Crash prediction models such as Tobit regression, exponential regression, Poisson, negative exponential model and cluster analysis, are assessed to look at easier access to entrances and exits into highways. These predictive modeling tools assess the factors causing crashes as well as where to target and prioritize future projects in terms of crash likelihood. Geographical Positioning Systems (GPS) enable a greater availability to acquire crash data such as the Crash Records Information System (CRIS) developed by the Texas Department of Transportation (TXDOT). The econometric methodology developed in this study shows that crash rates have a direct relationship with access density and flow. The model in this analysis provides a better understanding of the physical characteristics leading to the increase in crashes