Pinpointing crash ‘hot spots’ through mapping
Reducing crash-related injuries and fatalities is a major focus of every law enforcement agency and countless education campaigns, but the resources available for those efforts are finite. One question is how to use these limited resources to make the biggest impact.
“Building a better map to identify crash hot spots is essential,” said William Schneider, an assistant professor in the University of Akron Department of Civil Engineering and Roadway Safety Institute (RSI) researcher. “Law enforcement organizations and motorist education campaigns have a set amount of money, and optimizing hot-spot mapping for these groups allows them to better manage their resources and improves both education and enforcement.”
Traditionally, crash-mapping techniques have been either point-, segment-, or zonal-based. Each of these techniques has advantages and disadvantages, and often these basic statistical tools provide only preliminary results. During an RSI seminar, Schneider discussed his research that is focused on overcoming the disadvantages of these methods through advanced statistical techniques, increased computing power, and improvements in crash location technologies.
“By combining traditional crash-mapping techniques into a more advanced analysis, we can overcome many of the weaknesses of these basic analysis methods,” Schneider said. “For example, we are using the exact same crash data but joining both point- and zonal-based analysis to remove ambiguous zonal borders and improve the accuracy of crash hot-spot identification.”
Schneider’s research also addresses a number of other crash-mapping challenges by looking at clustering of high-frequency crashes, examining the relationships between crash clusters, calculating the statistical significance of crash clusters, and developing better models through estimation and testing. Current and future research projects will further refine these models by improving crash cluster identification, developing new methods to correlate multiple data sets, improving model estimation, and using multiple variables to pinpoint crash clusters.
In a current project for the Roadway Safety Institute, Schneider is building on this research to identify geospatial trends in alcohol-related crashes to help law enforcement target efforts for preventing them.
Ultimately, better crash mapping will not only lead to more targeted education and enforcement efforts, but also help improve long-term resource allocation for highway safety stakeholders.
“This information can guide decisions about long-term budgeting,” said Schneider. “It will help organizations manage their resource allocation to do more with less and offer a non-biased way to justify why those dollars are being spent in a certain location instead of somewhere else.”