In January, University of Florida researchers presented a report to the Florida Senate Agriculture Committee estimating that three hurricanes caused up to $975.8 million in ag production losses for 2024.
After a hurricane passes, data on how the storm affected local crops can be slow to come out.
Now, researchers at UF’s Institute of Food and Agricultural Sciences want to give farmers an artificial intelligence tool to get near-real-time results.
Nikos Tziolas, assistant professor and lead scientist on this for UF/IFAS, said it's a simple chat-based interface, which uses satellite imagery to answer questions conversationally.
The video below shows how the chat will function:
He said it will function a lot like ChatGPT.
"The user, the farmer, can open the chat and ask, ‘What are areas are flooded in my field?’ Or, ‘How does my crop health compare to last year after the event,’" he said.
"We have a cyber infrastructure [which] enables us to automatically download and process data from the whole Floridian territory," he continued.
And you don't have to be an expert to use this technology because it's so conversational.
Tziolas hopes this can also help agriculture extension agents in their line of work.
“The key component here is that users, non-experts in our case, like farmers or extension agents, have an opportunity to interact with the system using a simple natural language question,” he said.
Drones can always be used, but Tziolas argues that the method is usually more costly without the added benefit of utilizing maps and analyzing trends, like his AI model.
The prototype will soon be tested at select farms within Immokalee in Collier County and Ona in Hardee County.
Then the farmers can provide the team with feedback in the next six months.
It could be fully functioning and widely available by next year's hurricane season.
A $297,000 grant from the U.S. Department of Agriculture’s National Institute of Food and Agriculture is paying for the tech development.
