A Smarter, Safer Way to Monitor Our Pittsburgh Streets

It’s a Monday, and you’ve overslept. You skip breakfast and rush to your car, hoping to make it to work on time. But a notoriously long red light delays you. You’re not just frustrated – you’re late. 

For an engineer who specializes in transportation, ensuring others arrive to work on time is just one of their priorities. 

“When we talk about road transportation, we have three areas that shape our work,” said Aleksandar Stevanovic, associate professor of civil and environmental engineering at the University of Pittsburgh Swanson School of Engineering. “Safety is our primary concern, but we also think about efficiency and environmental implications. Most of the time, these areas go hand-in-hand.” 

Stevanovic has spent the majority of his career researching traffic control–which can be through road signs, traffic signals and pavement markings–on arterial streets. He has published more than 250 journal and conference papers and presented at more than 100 international, national and state seminars and professional meetings on the subject. He has been principal investigator on ~40 research projects for almost $5 million in funding, and has worked with various transportation agencies on the national, state and local levels. 

That’s a lot of time spent on the road(s). 

Smarter Streets 

Most recently, Stevanovic is part of a $160,814 project with the city of Pittsburgh to improve the city’s intelligent transportation system’s infrastructure and operations in disadvantaged neighborhoods, or SmartPGH. With Stevanovic’s help, the city plans to deploy different smart technologies on its streets that support multimodal operations. 

“By multimodal, I mean accessible public transportation by providing transit signal priority for buses, providing better detection of pedestrians, and more protected signal phasing for crossing the street,” Stevanovic said. “We are also looking into bike volumes and what areas people are delayed in their cars.” 

There is one caveat for Stevanovic and his team: when these technologies are installed, it’s uncertain how efficiently they’ll function together since they typically operate independently. Worst case scenario: less than ideal performance from these technologies.. 

“We can’t test in the field because we could potentially put lives at stake,” Stevanovic said. “Instead, we use a very sophisticated, high fidelity simulation to replicate the conditions that we have in the field with the same technologies that will be implemented.” 

The simulation is more similar to that of a digital twin than virtual reality, meaning the research group can control different conditions to produce a result in order to form accurate predictions. 

“You don't get like this 3D view of the objects around yourself, but the way how traffic performs, number of vehicles, how fast vehicles move, number of pedestrians, where pedestrians move, and how they move in the network,” Stevanovic said. “Those things are very realistic.” 

Reinventing the (Steering) Wheel 

Unlike SmartPGH, Stevanovic and his team are incorporating virtual reality into a $120,000 project, in collaboration with the University of Virginia, under a consortium of regional universities led by Morgan State University .

In his lab, there’s an interactive environment where one person drives a car, another walks as a pedestrian, and a third rides a bike—all within the same virtual space. All other cars, pedestrians, bicyclists and landscapes are simulated. The environment itself is based on a corridor in Newark, Delaware because of its notable multimodal environment, maintained by the Delaware Department of Transportation. 

Stevanovic emphasized that calling it a mere driving simulation system doesn't capture its complexity. 

“This creates a perfect environment for us to investigate multiple, multimodal traffic scenarios,” Stevanovic said. “By using these human agents in that environment, we can replicate some of the potential conflicts that happen and test the ability to avoid and resolve these conflicts without actually working with the people in the field. No one gets hurt.” 

Though remote  work has become more common, Stevanovic doesn’t believe it to be the perfect solution to traffic conditions. Stevanovic said it’s possible that multidisciplinary researchers could look at something as simple as the wheel – which dates back to the 4th millennium BC – and a car’s design to improve traffic conditions. 

“We’ve never really questioned what a car should be—its design, features, or purpose,” Stevanovic said. “We typically follow evolutionary changes, not revolutionary ones. Maybe it’s time for a radical redesign to improve how we use our time, reduce delays, enhance safety for everyone, and make cars more eco-friendly.”