Alcohol-Related Hot-Spot Analysis and Prediction

Abstract

This project developed methods to more accurately identify alcohol-related crash hot spots, ultimately allowing for more effective and efficient enforcement and safety campaigns. Advancements in accuracy came from improving the calculation of spatial autocorrelation and interpolation, the identification of spatio-temporal patterns, and the influence of geographical patterns on the spatial distribution of crashes. The project then used the location-based hot-spot maps created using these improved methods to develop a new method of patrolling for intoxicated drivers. This method guides officers to statistically significant locations where intoxicated drivers are most likely to be, allowing officers to be more accurate while patrolling. Additionally, this method allows officers to pass through more alcohol-related crash locations per minute and mile than current patrolling practices. By improving how officers patrol, individuals may be deterred from driving while intoxicated, and alcohol-related crashes may be ultimately reduced.