Is there a spot on your drive to work where you regularly have close calls with other drivers? Maybe it’s a highway on-ramp that feels too short, or an intersection where everyone blows through red lights. 

For urban geographers like Congcong Miao ’26 Ph.D.,  these spots aren’t random – they are what happens when urban planning fails to account for human behavior. And while, in the best-case scenario, they may raise your heart rate on your commute, they are also a contributor to car accidents. In 2025 alone, Connecticut saw nearly 100,000 car crashes. Two hundred and fifteen of these involved at least one fatality.

The good news is that, with enough data and dedicated research, these spots can be identified and fixed. They can even be prevented, with research-based intervention helping urban planners avoid constructing risky environments. 

Congcong Miao, who is graduating this spring with her Ph.D. in geography, got interested in this field when she was working as a research assistant for the Connecticut Transportation Institute (CTI) headquartered in UConn’s College of Engineering. 

“The first project I worked on at CTI was visualizing traffic crash data across the state of Connecticut since 2015,” recalls Miao. “I remember feeling quite surprised when I first saw the data. You know, my everyday experience driving on the road in Connecticut feels relatively safe. But when I saw the traffic crash data, I realized that actually, a substantial number of crashes occur every year. Of course, many of those crashes are not severe crashes, but the overall numbers were still higher than I had expected.” 

The statistics got her curious. Were there places with disproportionately high crash rates? And if so, what characteristics did those places share?  

This line of inquiry led to Miao’s first published paper as a doctoral student, in which she developed a spatial-temporal statistical approach to answer these questions. This paper introduced an important new methodology in the field. Rather than just tracking areas with high crash rates (a spatial analysis), it also emphasized the time at which these crashes occurred (temporal analysis), as well as the severity of the crashes.  

In her later work, Miao has also uncovered a surprisingly human dimension of road safety. She found that it is not only the physical characteristics of an environment that impact safety – it is also how people perceive these characteristics. 

My research is grounded in a simple idea: the environment can shape human safety both directly and indirectly.

Environments perceived as “beautiful” and “safe” create a positive self-fulfilling prophecy: people generally travel more safely in these spaces, and are less likely to get into accidents. On the flip side, environments perceived as “boring” or “depressing” tend to have more crashes. 

“My research is grounded in a simple idea: the environment can shape human safety both directly and indirectly,” says Miao. “Physical conditions — such as the road surface conditions, the surrounding buildings, or sun glare — can directly influence the likelihood of traffic crashes. But it’s not only about how the environment is built, it’s also about how people feel and how they experience the environment.” 

Healthy Habitats

Beyond just car crashes, Miao has also worked tirelessly to develop new methodologies to answer more questions about urban planning and its intersections with public health. 

Earlier this year, she was awarded the Peter Gould Paper Award from the American Association of Geographers (AAG) Health and Medical Geography Specialty Group. This award recognized her work using mobility data to understand how people access food – whether they are able to access healthy food, and how this impacts their personal health outcomes. 

She has also published research on how people are able to access healthcare from specific providers, such as pediatric dermatologists and Medicare-participating dermatologists. As with her research on food security, she has identified access challenges and policy recommendations to help solve them. 

“This work reflects my broader interest in leveraging large-scale geospatial data to study human behavior and inform more effective and equitable policies,” Miao says. 

Throughout her doctoral studies at UConn, Miao has worked and published with professors Xiang “Peter” Chen, Chuanrong “Cindy” Zhang, and Suining “Henry” He, who share her interests in spatial analytics and health geography. 

Tools of the Trade

Much of Miao’s work uses a combination of several types of sophisticated software tools, especially geospatial artificial intelligence technology, or geoAI. 

AI can be helpful in many layers of her research, she explains. For example, geoAI tools can help analyze images of urban environments, such as those from Google Street View, to determine their physical characteristics.  

“Using AI tools, we can extract detailed characteristics from images that can represent the environmental conditions of the city we’re living in,” says Miao. “And this allows us to systematically examine what kind of environmental conditions are associated with high risk, and/or what kind of environments can be more dangerous in regard to traffic safety, like with frequent traffic.” 

AI tools can also help researchers like Miao understand human behavior patterns. This is especially vital to help make sense of the highly complex, massive, noisy data sets geographers are often working with. 

This technology has also unlocked a new research frontier for Miao: helping refine AI tools to be more reliable and meaningful for geographic analysis and real-world applications. 

“Many of the AI tools we use today, they may be able to generate maps, but they may produce errors because they lack the basic geographic literacy,” she explains. “That gap is something I’m especially interested in as a researcher. How can we design AI systems that can better understand geographic concepts and spatial relationships in context, so they can develop a form of geographic intelligence?” 

Miao will be exploring these questions, among others, in the exciting next step of her research career: She has accepted an assistant professorship in spatial data science and geoAI at Binghamton University, where she will begin teaching and researching in the fall.