How Universities Are Turning Their Connected Campuses Into Smart Cities

University of Michigan Project Changes Course Due to Regulations

The University of Michigan’s Transportation Research Institute’s Mcity is an advanced mobility research center. As a public-private partnership among the university, the city of Ann Arbor and dozens of other government and industry entities, it’s been focused on connected, automated technologies that are the backbone of intelligent transportation.

While its work continues, regulatory shifts have forced a change.

“We stood up dedicated short-range communication, a 5.9-megahertz dedicated band. Then the FCC came along and truncated our band and said you cannot use dedicated short-range communications anymore,” says Debra Bezzina, managing director for the Ann Arbor Connected Environment.

“We had to recall all of the vehicles that we had deployed, and at that time, we had about 1,650 of them,” she says. “They all had to come back in and have their devices taken off. All 75 of the roadside units that were deployed in Ann Arbor had to be turned off.”

DISCOVER: Modern networking solutions can enable smart city research.

The research continues, but in a new form. “We have transitioned to a cellular-based solution: Cellular V2X, Vehicle to Everything. It’s still in the same band and it’s still direct communications, but this is based on cellular,” Bezzina says.

Funding has come in the form of U.S. Department of Transportation grants, including one for the Smart Intersections Project that delivered $10 million in federal funding and a $10 million cost share. 

“That was to equip 21 intersections with C-V2X technology and perception systems, a combination of video cameras and LIDAR,” she says. Another project, the Ann Arbor Connected Environment Reimagined (or ACE 2.0) got about $12.5 million, with $10 million in federal funding and a $2.5 million cost share.

“Then we have a third, smaller project, part of the city of Ann Arbor’s Safe Streets,” Bezzina says. Here, the team is developing a system that helps detect near misses in traffic situations, “and we’re going to be introducing an AI agent to help.”

When considering near misses, “you don’t want to only say, ‘If I’m 99% sure, I’m going to call that a near miss,’ because then you’re going to miss a lot of things. But you’ll also have a very low false-detection rate. So, it’s always balancing those two things,” she says.

“When you use AI, you can balance them better, and when you add an AI agent, you just go to that next level, to be able to say, ‘For these ones that have a lower confidence, we’ll have the AI agent actually go through there and identify them,’” Bezzina says.

Funding Drives a Pivot at Portland State University

Portland State University established its Digital City Testbed Center along with the University of British Columbia in Vancouver. It worked closely with Portland’s Bureau of Planning and Sustainability and the Portland Bureau of Transportation to explore things such as air quality sensors that monitor vehicle pollution.

That effort wound down in 2023. It was originally funded with a large grant from Portland State University, along with a $150,000 grant from the National Science Foundation. But follow-up funding proved hard to land. 

“We submitted three much larger proposals to NSF that got reasonably good reviews, but never were selected for full funding,” says Jonathan Fink, a professor of geology who headed up the Digital City Testbed Center.

The project demonstrated success. For example, researchers worked with the British company Waymap to develop a navigation system for blind people, which was eventually deployed throughout the entire Washington, D.C., area. “A blind person can use the app to get from point A to point B in the Washington, D.C., metro, with step-by-step instructions,” Fink says.

LEARN MORE: Hybrid infrastructure powers AI research projects in higher ed.

In a separate effort, the researchers have also explored linking Wi-Fi-based occupancy data to PSU’s heating and cooling system. An in-kind donation worth $750,000 from Cisco helped here, supporting a three-year license for Cisco’s most advanced wireless networking system. 

“We worked with Cisco and a startup called Sensible Building Science to control heating and cooling based on the number of people in a given room” based on use, Fink says. It was promising, but ultimately unsuccessful: The startup’s software wouldn’t integrate with PSU’s IT security protocols, so “we were not able to scale our initial results.”

When funding for the test bed effort dried up, Fink pivoted to a role facilitating new work around wildfires management. The researchers now are developing detection mechanisms, “and also looking at the cities where the smoke was affecting large populations,” he says.

“A lot of different technologies are being used to monitor where fires are occurring, to catch them at the earliest stages,” he says. “There are drone-based systems for either fighting fires or monitoring fires, or for thinning forests. There are also technologies for making urban areas less flammable.”

AI helps to move the work along. University researchers are collaborating with tech vendors and city officials, and as a facilitator of these conversations, “I’ve used AI to take the recordings of the webinars and turn those into a resource,” Fink says. “I’ve used Notebook LM to take all of the transcripts and digitize those, upload those and make those into a searchable tool.”

Researchers Reflect on Lessons Learned from Smart City Projects

How can you succeed with university-based smart cities research? Those who have been there point to a few key lessons learned.

It’s important to be proactive in managing the relationship between the school and the city. To that end, Bezzina looks to build enthusiasm among city leadership. “You really can’t just deploy this stuff and say, ‘It’s good for you,’” she says. “You have to address real-world problems, adding applications that are important to them.”

The city of Ann Arbor, for example, has issues around pedestrians and bicyclists who interact with buses. “We can use this technology to resolve it,” Bezzina says, and she’s worked hard to communicate that benefit to the city. “You have to understand, they have other priorities. You can’t be breathing down their necks and demanding things.”

Experts say this approach makes sense. In a successful effort, the university “brings in local government officials to see firsthand how it works and learn about the benefits of implementing new smart city technologies,” says IEEE Fellow Karen Panetta.

It also makes sense for schools to be thoughtful when selecting the technology partners who will support their smart cities research. Some of Fink’s work, for example, has relied heavily on startups, and that can be a gamble, he says.

“We were really depending on these startup companies to come up with ideas that we would scale on the various campuses. Several of those just didn’t work out,” Fink says. He adds that his team has been able to offset this risk through partnerships with big players such as Cisco and Microsoft — a key learning for schools looking to ramp up their own smart cities research.