A two-step geo-grid screening process to identify locations for safety improvements

MAP-21 emphasizes highway safety through programs that target reducing crashes as demonstrated by an 88% increase in funding for the highway safety improvement program (HSIP) from 2012 to 2013. This translates to a median 2013 HSIP funding level of just $33.6 million at the state transportation agency level. Given the high cost of major facility improvements, these funds would potentially only address improvements for a small percentage of the “top” crash locations in a state. Having the ability to quickly identify multiple high crash locations that resulted from factors that can be addressed by lower-cost, readily implementable solutions, could substantially improve the return on investment for safety improvements. This study develops a two-step geo-grid screening procedure using a simple to use web-based GIS tool which enables anyone with access to crash data, including new hires with little or no local knowledge or GIS capabilities, to quickly identify high crash locations where viable safety improvement options exist. The area under consideration is initially divided into a screening grid which is used to aggregate crashes that are filtered by characteristics of interest. Cells with the highest number of crashes are highlighted allowing users to quickly identify focus areas. Users then select any cell in the screening layer, and the process is repeated on newly generated localized cells that are small enough for pinpointing specific crash locations. Engineers can then review individual crashes, local roadway geometry, traffic patterns and associated traffic control mechanisms to determine possible causes and propose improvement options for mitigating them. Building on a simple yet practical method for estimating the economic losses due to roadway crashes developed by the authors, an alternative to aggregating number of crashes was also implemented where each crash was assigned a dollar value based on its severity level. Total comprehensive crash costs were then used in place of crash counts for identifying high-dollar cells. Both approaches were tested using Virginia crash data to identify potential locations for safety improvements. The results provide useful information to support resource allocation.