PROVO — A new artificial intelligence program is helping Utah’s mosquito control crews discover thousands of storm drains they didn’t know existed, giving them new tools to prevent mosquito-borne diseases before they become problems.
The technology, developed by the Utah County Information Systems Department, analyzed high-resolution aerial photography and identified approximately 25,000 or more storm drains not included in the abatement team’s existing database.
These newly identified drainage channels are important as they can become major breeding grounds for mosquitoes.
“What we’re trying to do is prevent West Nile virus and all kinds of mosquito-borne diseases from occurring in our county,” said Utah County Mosquito Control Director Jason Byrd.
Mosquito control crews on bicycles treat the county’s storm drains throughout the spring and summer with water-soluble packets designed to kill mosquito eggs and larvae before they hatch.
“Certain mosquitoes love storm drains,” Bird says.
One of those helping with the effort, Alyssa Collins, rides her bike from drain to drain, using digital maps to pinpoint treatment locations.
“The map already has all the storm drains marked. Everything is marked,” Collins said. “You just ride your bike to where there’s a little mark and throw it in there.”
Finding all the stormwater drains used to be one of the biggest challenges.
Utah County believed it had approximately 50,000 storm drains mapped within the county.
As the county continued to grow, officials recognized that records were likely to be incomplete.
“We didn’t know exactly where all of them were,” Bird said. “I think the population has increased about 40 percent since 2010 because of the growth of Utah County.”
That’s when his team asked the county’s information systems team to leverage artificial intelligence.
Utah County Information Systems Director Patrick Wallo said GIS programmer Nathan Thomas developed a model that can recognize storm drains in aerial photography.
“He designed an AI model that examines our latest high-resolution aerial imagery, and we train the model to learn what to look for,” Walo says. “We taught them what a storm drain looks like. Sometimes a storm drain is round, sometimes it’s square, sometimes it’s rectangular. But this is what a storm drain looks like, and you learn by doing a few tests.”
AI scanned images across the county and identified thousands of additional storm drains.
“It opened our eyes to the fact that there are actually a lot more storm drains out there,” Byrd said.
County officials say the technology helps update outdated GIS data, allowing mosquito control crews to spend more time treating drains instead of searching for them.
“This allows us to detect the location of storm drains without the technician having to spend time trying to find them,” Bird says.
Finding an additional 25,000 stormwater drains will mean more work for mosquito control crews, but officials say it’s a job well worth doing.
“What we’re trying to do is eliminate the problem before it becomes a problem,” Bird said. “Success in mosquito eradication is when nothing happens. No headlines, no outbreaks. That’s success for us.”
