A $1M Project to Prevent Wrong-Way Driving

Ten years ago at a meeting near Ames, Iowa, a police officer told officials from the state Department of Transportation that his dispatchers were being inundated with 911 calls about drivers going in the wrong direction on U.S. Highway 30.

“He said, ‘I don’t know how many I’m supposed to be getting, but it seems high,’” said Willy Sorenson, a traffic and safety engineer for the Iowa DOT. “And then he asked if we could help.”

Sorenson, who often spearheads special projects for the state agency, was intrigued by the problem. Wrong-way driving—where cars travel in the wrong direction on one-way roads—is a common problem on highways across the country, but officials typically only know about incidents when they result in a collision. That makes it difficult to prevent or track the problem before crashes occur. But maybe emerging technologies could change that, Sorenson thought.

“I said, ‘We’ll figure it out,’” Sorenson said. “It seemed fitting to figure out how to use technology to help us collect data, and once we did that, we could formalize a plan. Because without data, how do you tackle something in a wise way?”

Sorenson began by accessing 911 call data from the area to confirm both the volume and general location of reports of wrong-way driving, which occurred most often on a 25-mile stretch of highway between the cities of Nevada and Boone. For four years, Sorenson attempted to address the problem through what he calls “low-cost treatments”—primarily increased signage near on- and 0ff-ramps, including reflective placards reading “wrong way,” “one way” and “do not enter.”

“Over the years sometimes they get hit or run over, or they’re faded, so we replaced those and put up a couple of extras here and there,” he said. “But because we were only measuring it using those 911 calls, it was hard to see results.”

Emergency calls are not always the most reliable measure of wrong-way driving, Sorenson explained, because many errant drivers quickly recognize their mistake and self-correct their bearings before anyone sees or reports them. But because wrong-way driving crashes have higher fatality rates than other collisions—1.34 deaths per incident, versus 1.1 for all other car accidents—the problem seemed to warrant further research.

“We get about 20 wrong-way driving crashes per year, out of about 55,000 total crashes,” Sorenson said. “What makes it horrendous is that the chances that somebody dies is so much higher. Whether it’s the angle of the crash, the unexpectedness of seeing someone coming in the wrong direction, you don’t really have that same chance of swerving out of the way.”

Four years into Sorenson’s research, the Iowa DOT announced plans to install cameras and traffic sensors along the same stretch of highway he was studying. He asked if some of those cameras could be aimed at on-ramps to help detect cars going the wrong way, a request the department declined—but officials did offer to configure the sensors in their planned locations to be able to detect wrong-way drivers. 

Most of those locations were at least a mile from points of entry, which was less than ideal for determining exactly where and how drivers were becoming confused, Sorenson said. But the sensors did provide some useful information, including clear indicators that most wrong-way driving incidents are occurring at intersections, rather than on-ramps.

“We’ve collected information on 232 drivers with these sensors,” he said. “We’re also working with law enforcement, so every time they get a 911 call about a wrong-way driver, they send it to me, too. Then I send it to my interns, and we do a little contact-tracing, in a way, by looking back at our camera footage to see if we can pinpoint where each car started.”

Sorenson assembled that information and presented it to his superiors, showing what his team was able to do with relatively limited investment. After that, the department appropriated funding for two additional projects—$1.3 million in federal dollars for enhanced signs at 165 intersections across the state, and $300,000 in state funds for 60 new cameras to help pinpoint starting locations for wrong-way drivers.

To determine the best locations for the signs, Sorenson ranked each of Iowa’s 467 interchanges based on a number of factors, including traffic volume, crash frequency, 911 calls and each intersection’s proximity to bars and taverns.

“We’ve found that about 60% of wrong-way driving is alcohol related,” Sorenson said. “You don’t have to drink at a bar to go the wrong way, but it certainly helps.”

The highest-risk locations will also be outfitted with cameras that have the capability to detect a wrong-way driver and alert staff via email (about 50 have been installed so far). Those messages are analyzed by staff members to determine whether the enhanced signage is working, though Sorenson may at some point explore the possibility of having the alerts sent to law enforcement as well.

“We’re in the phase of configuring the camera to filter out the false calls,” he said. “They aren’t perfect. If a heavy cloud goes the wrong way, for example, and casts a shadow on the pavement, the camera thinks it’s a car. Once we’re out of that phase our goal is to notify our traffic management center with these alerts so they can confirm them and then call 911 if necessary.”

The enhanced technology allows Sorenson’s team to gather more detailed information about each car that drives the wrong way, including the path the car took, where it traveled, how far it got and whether the driver self-corrected without outside intervention. That information is valuable in building preventive measures that are uniquely tailored to each problematic intersection, he said.

“I want to solve wrong-way driving before it happens, which means a lot of our treatments are proactive,” he said. “We’re doing things to the roadway system so that when drivers approach where they could go the wrong way, there’s better signage so they don’t do it to begin with. And that in itself is a little bit different. Historically, we just kind of threw out signs and assumed that people would see them. But now we have the technology to confirm whether that’s true.”