Mangos. Sweet, juicy, and delectable—that is, at the peak of ripeness. If picked too early, they arrive in grocery stores green and hard; if picked too late, they arrive damaged or spoiled. To maximize yields and earn top dollar for their crops, growers must time their harvests carefully, a process that generally relies upon imprecise visual inspection of the fruit.
Three young women who attend the Federal University of Pernambuco in Recife, Brazil have come up with a better way. As participants in a 2014 Intel Embedded Systems Competition for university students in Brazil, Camila, Vanessa, and Vanessa, developed “QualisFruit*,” an Intel® Atom™ processor-based system that uses a matrix of sensors to detect and count yellow pixels of color in mangos to precisely determine their stage of ripeness.
They used a very creative way to solve the mango selection issue, using already known techniques in a specific way that nobody thought of before.
Rubem Saldanha, Intel Education manager
“We wanted to develop a project that promotes some benefit to our state, and agriculture is strong here,” says Vanessa. In fact, she adds, 90% of the mangos exported from Brazil are grown in the dense orchards covering some 57,000 acres in the Petrolina area.
The young women hope to turn their system into a commercial product. “Our market in the future can be growers and fruit-packing companies. We are testing the current system version in a controlled way, but soon we will initiate field tests,” says Vanessa, noting that they also plan to adapt the project to analyze other types of fruit.
The team captured the top women group award in the Intel Embedded Systems Competition, which is held annually in partnership with the Brazilian Computer Society. “It’s important for students to realize that they can think about real problems of the world and use the subject they are studying at the university to solve these problems,” says Saldanha.
This team’s confidence about the future of their project as a product could solve an issue for the mango industry.