A new study from Carnegie Mellon and UT Dallas suggests self-driving cars could ease parking headaches but also increase traffic, pushing cities to rethink fees, curb space and downtown land use. The findings offer a roadmap for urban planners preparing for an autonomous future.
Autonomous vehicles may someday spare commuters the daily hunt for a parking spot downtown. But a new study suggests that without smart policies, self-driving cars could also mean more traffic on the road and new challenges for cities.
Researchers from Carnegie Mellon University and the University of Texas at Dallas used Pittsburgh as a case study to explore how autonomous vehicles, or AVs, might reshape morning commutes and parking in central business districts. Their work, published in the journal Management Science, focuses on what happens when AVs become the default way people drive to work.
The team’s message to city leaders is clear: the technology is coming, and the rules of the road need to catch up.
“Urban planners have a rare window of opportunity to establish policies that pave the way for the inevitable mass arrival of AVs,” co-author Soo-Haeng Cho, the IBM Professor of Operations Management and Strategy at Carnegie Mellon’s Tepper School of Business, said in a news release, .
Today, downtowns devote huge amounts of land and money to parking structures, yet many commuters still face high fees and limited options near their offices. At the same time, traffic congestion during the morning rush adds time, stress and pollution to daily travel.
AVs could change that equation. Because they can drive and park themselves, self-driving cars could drop commuters at their office doors, then head to cheaper parking lots in outlying neighborhoods or suburbs. That could save drivers money and reduce the need for large, expensive parking garages in the urban core.
To understand the tradeoffs, the researchers built a continuous-time, game-theoretic traffic model of the morning commute into a central business district. In simple terms, they created a mathematical simulation of thousands of commuters making individual choices about when to leave home and where to park, while responding to real-world factors such as parking prices, congestion, and curbside pickup and drop-off rules.
The model compared a world of human-driven cars with a world in which commuters all use AVs.
One key finding: when everyone has access to AVs, many commuters are likely to choose cheaper parking outside the central business district and let their cars drive the extra distance on their own. That behavior would increase both vehicle hours and vehicle miles traveled compared with human-driven vehicles.
In other words, the cars would spend more time and distance on the road, even if people spend less time behind the wheel. The study concludes that this shift would raise the total cost to the transportation system as a whole, including congestion and infrastructure wear and tear.
At the same time, the model suggests that business districts could reclaim valuable land now tied up in parking. If AVs park themselves elsewhere, downtown parking lots and garages could be repurposed for housing, offices, retail or public spaces, potentially making city centers more vibrant and walkable.
The researchers emphasize that their work is not a blueprint for one city, but a way to understand general patterns that AVs may trigger in many places.
Neda Mirzaeian, an assistant professor of operations management at UT Dallas’s Jindal School of Management who led the study, added, “In our study, we sought not to propose city-specific solutions, but to highlight general tradeoffs and dynamics in human behavior that emerge when AVs, commuters, and infrastructure interact,” underscoring that the same forces could play out in different ways depending on local policies and geography.
The study also points to tools that city officials can use to steer those dynamics in a better direction.
According to the model, urban planners could reduce the overall system cost by adjusting parking fees, imposing congestion tolls during peak hours, or redesigning infrastructure. For example, some existing parking spaces in central business districts could be converted into designated drop-off and pick-up zones for AVs, smoothing traffic flow and reducing curbside chaos.
In the Pittsburgh case study, the researchers estimate that such measures could cut total system costs by up to 28.5%. That figure reflects the combined impact of less congestion, more efficient use of road space and smarter land use, rather than just savings for individual drivers.
Mirzaeian noted that the model is meant to help cities anticipate the ripple effects of policy changes in an era of rapid technological change.
“Our model can serve as a guide, or even an early warning system, to recognize how seemingly small shifts in technology, costs, or incentives can lead to large changes in commuter behavior and system-wide efficiency,” she said.
The work also speaks directly to the growing number of public agencies now grappling with how to regulate AV testing and deployment on their streets.
“In providing guidance to urban planners—including mobility and infrastructure departments of mayoralties, city councils, town councils, and town boards—our results can identify when and where current policies need to adapt in light of the special needs and characteristics of AVs when AVs become widely deployed,” added co-author Sean Qian, the H. J. Heinz III Professor of Civil and Environmental Engineering at Carnegie Mellon’s College of Engineering and Heinz College.
For students and young professionals interested in transportation, urban planning or public policy, the study highlights how math, economics and engineering intersect with everyday life. Commutes, parking prices and curb space might seem like mundane details, but they add up to big questions about how cities grow and who benefits from new technologies.
As AVs move from pilot projects to mainstream use, the researchers argue that the choices cities make now about pricing, zoning and street design will determine whether self-driving cars simply add more vehicles to already crowded roads, or help build more efficient, livable urban centers.
Their conclusion is less about the cars themselves and more about the rules that will govern them: autonomous vehicles will not automatically solve the commuter parking problem, but with careful planning, they could help cities rethink it.
Source: Carnegie Mellon University