Dedman College of Humanities and Sciences Dedman College Research Faculty News Statistical Science

Placing neighborhoods in focus

SMU News

Researchers combined street-level investigations with SMU’s supercomputer power to reveal infrastructure deserts. Their study lays the groundwork for improving neighborhoods.

Residents of a neglected corner of southeast Dallas daily navigate crumbling sidewalks, pothole-riddled streets and neglected intersections. Few trees shade their streets, and the lack of access to basic services like internet, health care and grocery stores isolate them within a thriving city. Like residents of 61 other Dallas neighborhoods, they live in an infrastructure desert.

What are infrastructure deserts? Why do they matter?

Those two questions get to the heart of a multiyear research project led by SMU’s Barbara Minsker, a nationally recognized expert in environmental and infrastructure systems analysis.

To find answers, Zheng Li, a Ph.D. student in civil and environmental engineering, and other team members created a computer framework with the ability to assess, at census-block level, 12 types of infrastructure. Neighborhoods were evaluated and compared by infrastructure deficiency, household income and ethnicity.

“This framework enables us to collect data from a huge variety of sources, then analyze the patterns that emerge to discover new information that can be used by scientists, policymakers and residents to improve their neighborhoods,” Li says.

More than 20 undergraduate and graduate students, faculty, staff and community members contributed to the research.

The multidisciplinary team included co-authors Janille Smith-Colin, assistant professor, and Jessie Zarazaga, clinical associate professor and Hunt Institute Fellow, both from the civil and environmental engineering department in the Lyle School of Engineering; and Xinlei Wang, statistical science professor in Dedman College of Humanities and Sciences.

SMU’s supercomputer was put to work mining huge public datasets, including 5 terabytes of images from the city of Dallas.

Undergraduate and graduate researchers used drones, smart phone applications and artificial intelligence, as well as their own feet-on-the-ground observations, to gather information supplementing the public data. They also held brainstorming sessions to learn what residents felt were their communities’ greatest infrastructure needs. READ MORE