OVERLAND PARK, Kan. — The next time it rains in Johnson County, Kansas, the county’s stormwater engineers may be able to predict potential flooding hazards in advance thanks to a program it's testing that relies on artificial intelligence.
“It’s super exciting,” said Sara Smith, senior stormwater engineer. “I think it could be really useful in the future by getting out ahead of the storm. Instead of being reactive, we can be proactive.”
The program is called Neer. It uses forecasts from the National Weather Service and maps of the county to predict possible flooding risks up to 24 hours in advance.
After a storm event, Neer crunches data from the county’s existing “Stormwatch” system to compare how its predictions measured up against reality.
The software uses machine learning, a type of artificial intelligence, to teach itself how to make better predictions before the next storm event.
“Flooding is on the forefront of everyone’s mind. Nobody wants their home to flood,” Smith said.
She said the Neer program is currently active in the northeast section of the county. She hopes to enhance and expand it countywide soon.
Her goal is that eventually municipalities can rely on the system to block roads and evacuate homes hours before floods arrive.
So far, the county has spent about $40,000 with Neer. The county has in place a special 1/10th of one percent sales tax to fund stormwater management.
The tax collected more than $15 million in 2020.
