New hurricane recovery technology being tested on Florida roads during peak season

By Joey Garcia, University Communications and Marketing

As Florida enters the peak of hurricane season, Sarasota County transportation engineer Wafa Mahmoud is prepared for a worst-case scenario. For the past decade, she has joined teams driving county roads after hurricanes, documenting flooded streets, damaged traffic signals and other hazards to determine where help is needed most.

"Sarasota County experienced significant impacts from Hurricane Milton, and our teams faced the enormous task of assessing damage across hundreds of miles of roadway," said Mahmoud, who is also a USF doctoral student. "The quicker we can identify hazards and infrastructure damage, the better we can direct resources where they are needed most and support our community's recovery."

USF Assistant Professor Hao Zhou, a member of the Center for Urban Transportation and Research, wants to make that process faster using artificial intelligence.

Working with transportation professionals across Florida, including those at the Florida Department of Transportation and Sarasota County, Zhou is developing AI tools to help transportation officials assess storm damage more efficiently, improve traffic signal restoration and better understand transportation needs after hurricanes. The first technology, an AI-powered dashcam, is expected to begin testing later this fall.

"Our team covers a large roadway network, and after a major storm there can be a lot of information to collect in a short amount of time," said Paula Wiggins, interim senior transportation manager with Sarasota County Transportation. "This collaboration gives us an opportunity to test whether this technology can help staff document roadway conditions more efficiently and support the decisions we have to make during recovery."

USING AI DASHCAMS FOR POST-HURRICANE DAMAGE ASSESSMENTS

In the days after a major storm, transportation departments need a clear picture of the damage. Assessing roadway conditions across large areas requires crews to document damage to roads, traffic signals and other transportation infrastructure, often through a combination of drive-through surveys and field inspections. The process can be time-consuming and difficult to scale with limited staff.

Zhou believes one solution may already be sitting on the windshield.

"Road inspections are already done from inside a vehicle, looking through the windshield," Zhou said. "A dashcam is a natural extra set of eyes. By adding AI, it can automatically identify damage and even generate reports to support recovery efforts."

The AI-powered dashcam records inspection routes, creating a searchable visual database that crews can use to verify damage, generate photos for reports and identify missing road signs that may have been blown away during a hurricane.

Mahmoud helped develop the technology by providing roadway damage images and inspection reports from previous storms, allowing the AI to learn how transportation engineers document and classify roadway damage in the field before it is tested during routine maintenance. She believes the technology could significantly reduce the time inspectors spend documenting storm damage.

"After the storm, internet service is often unavailable in some areas, so inspection data isn't uploaded until teams return to the office," Mahmoud said. "That delay postpones assigning work to maintenance crews and beginning repairs. The AI dashcam will save time and reduce manual data entry by automatically filling out the entire inspection report."

cREATING EFFECIENT REPAIR ROUTES

Once damage has been identified, agencies face another challenge: Deciding where to send repair crews and equipment. Contractor crews responding after recent hurricanes often lacked tools to coordinate where they should go next. As a result, teams sometimes traveled back and forth between intersections instead of following the most efficient routes, extending restoration times.

"We're developing scheduling and routing tools that can tell crews which intersections to repair next and the most efficient way to get there," Zhou said. "The faster crews can restore traffic signals, the faster communities can travel safely after a storm."

Mahmoud sees similar opportunities to improve roadway inspections. Today, Sarasota County crews rely on printed maps and manually drive every road within a designated area, often doubling back through neighborhoods to ensure nothing is missed.

"The system will provide shorter routes that cover the entire roadway network without missing any segments," Mahmoud said. "Teams will know their assigned route in advance, along with the estimated driving time, and the system can automatically adjust for road conditions and closures."

TURNING SOCIAL MEDIA INTO ACTIONABLE INFORMATION

Not every transportation problem appears on a roadway camera. Residents often share information online about issues affecting their communities, whether it's flooding on neighborhood streets, blocked access routes or shortages that make travel difficult. Those observations can provide valuable context that traditional transportation sensors may miss.

"People often report problems that traditional sensors can't detect," Zhou said. "Social media can help agencies understand mobility needs and transportation challenges that might otherwise go unnoticed."

The project grew from the realization that residents often become on-the-ground observers during disasters. Posts about flooded neighborhood roads, fallen trees blocking exits or gas stations running out of fuel can reveal problems that roadway cameras and infrastructure sensors never capture. Zhou is using AI to identify those posts in real time, helping agencies spot emerging issues, better understand public concerns and respond more quickly.

Looking ahead, Zhou hopes to develop dashboards that automatically identify roadway incidents and flag urgent mobility concerns as they emerge.

"Social media gives agencies another source of real-time information," Zhou said. "If we can automatically identify those reports, we can help transportation agencies understand public needs and respond more quickly."

Ultimately, Zhou envisions these technologies being used together to help transportation officials quickly respond after hurricanes.