Building a Better Berry with Artificial Intelligence
Volunteer tasting panels have long been relied upon to guide decisions on which strawberries taste and smell best, and therefore should be bred for Florida's $300 billion crop.
Tasting panels and breeding for better taste have been around since agriculture began, and now a researchers at University of Florida are using a new approach to creating better berries. Artificial intelligence is now helping identify cultivars most likely to produce a berry consumers will love. Already having collected data from hundreds and hundreds of volunteer tasters, computers now help match the chemical component that creates aromas of sweetness and “strawberriness.”
“If we want to say which of these aroma chemicals in the strawberry are actually helping boost the sweetness, our older statistical methods have a really hard time doing that but machine learning has an advantage,” said strawberry breeder Dr. Vance Whitaker.
He says knowing what chemicals to breed for makes it more clear what his team, and Florida farmers should plant.
Earlier this year Dr. Whitaker's team introduced Florida farmers to the white strawberry, which is gaining in popularity because of its sweet flavor and unique appearance.