If you’ve ever been unlucky enough to experience a car crash, you know that after the dust settles and the smoke clears, there’s paperwork. The police officer at the scene fills out a standard form that includes check boxes for crash information, a place for a description of the accident as told by the driver(s) and a diagram to indicate how the accident happened. In the end, that paperwork gets turned into crash data that lives in a variety of databases for future analysis.
As transportation engineers and planners, our team at Gresham Smith uses crash data to identify high crash locations on our roadways. Whether we’re designing a roadway project or conducting a traffic operations study, we look for possible trends so that we can make recommendations that will improve safety for the surrounding community. However, since crash reports are initially written by a person, human error is inevitable. Mistakes might include indicating directions incorrectly, reporting the GPS coordinates to where the report was taken instead of where the crash actually occurred or mis-classifying the crash.
In Florida, there are two commons sources of crash data: FIRES, a repository of crash reports completed by law enforcement agencies, and Signal4Analytics, a system developed by the University of Florida that is updated daily. Additionally, the Florida Department of Transportation (FDOT) takes data from these sources and verifies it, compiling it into the Crash Analysis and Reporting (CARS) database.
When our team analyzes crash data, we use the verified CARS data and supplement with data from Signal4Analytics. The data is available in spreadsheet form, but rather than jumping into sorting numbers and running calculations we review the actual crash reports to first determine if the crash was coded correctly. In order to develop realistic, reliable recommendations, we need to know the whole picture.
For example, a crash may be coded as an “angle crash,” but by reading the narrative and looking at the crash sketch we can quickly determine that the crash actually involved a left-turning vehicle that was struck by a vehicle traveling in the opposite direction. Or a crash might be coded as a “sideswipe,” but by digging deeper we may learn that the incident was actually caused by the other vehicle swerving to avoid an object in the road. By understanding what actually occurred at the time of the crash, we can better determine if there is a crash trend at a location or if the crashes are truly random.
Once we verify the crash data, analyze the numbers and identify a trend at a crash location, we conduct a benefit-cost analysis to determine if an improvement project at that location would be worthwhile. The outcome of the analysis can make a difference in whether a project gets funded by taxpayer dollars and where it sits on the priority list, so it’s important that we’re working with accurate information.
A recent example of this process at work is the Carson Drive intersection on US 41 in Land O’ Lakes, Fla. The intersection sees many rear end and left-turning crashes, as well as crashes involving bicyclists. By evaluating the crash diagrams in detail and evaluating the site, our team developed short-term and mid-term cost effective improvements, including high-emphasis crosswalk markings, additional pavement markings and signage, reallocated intersection lanes, and modified median opening locations and types. By digging into the crash data we were better able to understand the whole picture and develop improvements that will improve the area’s safety and, in turn, reduce crashes.
As transportation engineers and planners, we have a responsibility to encourage safe environments that support all modes of transportation in our communities. By delving deeper into crash data and understanding the context in which reports are filed, we can make better informed decisions when it comes to recommending roadway safety improvements.