Researchers at Oregon State University have created a smartphone app that uses artificial intelligence to predict when avocados are ripe and ready to eat. This is significant because a lot of avocados go to waste due to overripeness.
Luyao Ma, an assistant professor leading the study, explains, “Our aim is to help consumers and stores make better choices about when to use or sell avocados.”
The app was developed using over 1,400 smartphone images of Hass avocados. It can predict firmness—with an impressive accuracy of nearly 92%—and internal quality, distinguishing fresh avocados from rotten ones with over 84% accuracy. This tech could easily adapt to help determine the quality of other foods, too.
The ideal outcome? An easy-to-use tool for consumers that helps them avoid the disappointment of cutting into an avocado only to find it has brown spots.
Moreover, this tool can streamline processes in avocado processing plants. For example, if the app identifies a batch of ripe avocados, those could be sent to nearby retailers first.
This advancement stands out from past methods, which relied on older techniques like manual feature selection. “We used deep learning to automatically gather more information about the avocados,” says doctoral student In-Hwan Lee, who collaborated on this project.
Ma had personal reasons for choosing avocados. As someone who enjoys avocado toast, she often felt frustrated by not knowing when they were perfectly ripe.
This research also addresses a huge global issue: food waste. According to the United Nations, around one-third of all food produced ends up in the trash. In the U.S., the USDA and EPA have set a goal to cut food waste in half by 2030.
Ma emphasizes that avocados are just the start. “This technology has the potential to help consumers and retailers across many types of food, making smarter decisions and minimizing waste.”
The findings were recently published in Current Research in Food Science, offering a valuable glance into how tech can help reduce food waste. For more details, you can check out the full article here.
