Skip to Main Content

Researchers in civil and environmental engineering are working with state transportation agencies to model and analyze the impact of work zones on automotive accidents to inform policies and practices to decrease safety risk.

Zhuoran Zhang, Burcu Akinci, and Sean Qian have created a work zone safety analysis model that proves the causal relationship between work zones and increased car crashes. The team’s model was rigorously developed and informed by four years of data on thousands of work zones from the Pennsylvania Department of Transportation.

“There are tons of studies on work zones and crashes, but the main issue is that a lot of them use associational analysis,” said Qian, a professor of civil and environmental engineering and director of the Mobility Data Analytics Center (MAC). “They’ve compared accident rates in areas with work zones to accident rates from areas without, correlating increased accident risk with the presence of work zones.

However, correlation doesn’t necessarily imply causation. Increased crashes during work zones does not necessarily mean work zones caused those crashes. In fact, crashes may be caused by weather conditions, aggressive driving behavior, excessive traffic speed, or roadway design, among many factors, which could be related to why and how often a work zone is set up. Zhang, Qian, and Akinci set out to demonstrate a conclusive link between the two, breaking down the factors that affect risk around work zones.

Many such factors were identified, including weather, road and pavement conditions, driving behavior, time of day, duration of work zones, and work zone configurations. While these can all be factors in accidents, not all of them are necessarily related to work zones. Other factors may shift over longer time scales, such as driver behavior or technologies for driving assistance. In order to limit potential for confounding factors, the team tightened the focus of their model to weekly increments. Their approach was to analyze a given road segment in the weeks before road work began, the week in which the work zone was operating, and in the weeks after the work zone had been removed. By comparing a great number of road segments of various attributes over a long time period and considering potential factors that influence safety, statistical causal relationships between work zones and crash risks were obtained.

Causal models are crucial for improving road safety around work zones because they help us identify and analyze the various factors that impact risk in each scenario.

Burcu Akinci, Department Head, Civil and Environmental Engineering

The findings confirmed that the presence of work zones can lead to additional crashes in general. However, this effect varies by work zone types and other environmental factors. For instance, they determined that work zones operating during the night do not increase the risk for accidents as much as those operating during the day, possibly as a result of additional work zone configuration requirements during the night. Likewise, work zone impact on crashes is more pronounced on work zones with long-distance and high-traffic volumes. This implies that there is a great need to improve safety for long and traffic-heavy work zones during the day time.

“Causal models are crucial for improving road safety around work zones because they help us identify and analyze the various factors that impact risk in each scenario,” said Akinci, head of the Department of Civil and Environmental Engineering.

There are still many routes of exploration open for the team in understanding the relationship between work zones and increased risk. They’d like to extend their model to other areas, forging relationships with state and federal transportation agencies to demonstrate and improve the robustness of their model.

They’d also like to build on this work by analyzing factors not directly addressed previously, such as increasing the time span of their study period to cover work zones lasting a duration of more than a week, or further analyzing the risk impact of different types of work zone. Qian, Akinci, and Zhang, a Ph.D. student in CEE, intend to continue providing infrastructure managers and stakeholders with research that identifies the risks for drivers posed by work zones and that provides solutions that can help decrease that risk.

This work was sponsored by the Mobility21, a US Department of Transportation University Research Center, and the Pennsylvania Department of Community & Economic Development.