Computer Program Predicts Crop Pest Outbreaks
Their foe is small, but formidable. That's why researchers in Taiwan are pitting artificial intelligence against the oriental fruit fly, a crop pest that can cause billions of dollars in damage to 230 types of fruits and vegetables in Southeast Asia, Hawaii, California and Florida. On Aug. 8, California issued an emergency statewide alert in response to catching 13 flies in the state, the New Scientist reported.
The new Taiwanese fly-spotting system automatically sends researchers data from traps set up in farms all around the island nation. The traps send in data every 30 minutes. Before the new system was invented, researchers checked fly traps manually every 10 days, the New Scientist reported.
A program researchers wrote analyzes data from the traps, taking into account how weather affects fly breeding and any given area's usual population of flies. The program is able to learn from experience and is able to predict outbreaks 88 percent of the time.
Cheng-Long Chuang of National Taiwan University, who is leading the research, told the New Scientist he expects his system to reduce oriental fruit fly damage in fruit orchards by 50 percent. Meanwhile, a guava orchard owner taking part in Chuang's study said he liked knowing the real-time status of his orchard. The system also reduces his pesticide use, he said.
Source: New Scientist