For centuries, the coffee industry has relied on the time-consuming process of cupping to determine the quality and taste of coffee beans grown around the world.
Cupping is a traditional tasting method that has become standard for producers and buyers in the sector. To check the quality of a batch of coffee, professional cuppers conduct in-depth visual analyses and taste tests that measure body, sweetness, acidity, flavor and aftertaste.
Because it is usually conducted so far down the supply chain, farmers, traders and roasters often remain in the dark about their beans.
Until now, that is.
Demetria, an Israeli-Colombian startup specializing in agricultural and food technologies, has launched a pioneering artificial intelligence-powered platform and near-infrared (NIR) sensor system. In a first for the industry, the technology can determine the taste and quality of green coffee beans at any stage of the supply chain.
“The coffee industry has been lacking new technologies,” Eduardo Shoval, co-founder and executive chairman of Demetria, told The Media Line. “In the last decade coffee has started some kind of process we can call ‘wineification.’ People really want good coffee.”
Demetria, dually headquartered in Bogota and Tel Aviv, recently successfully completed a pilot with Carcafé, a Colombian coffee division and one of the world’s largest coffee trading houses. Working together with Carcafé, Demetria developed a solution that includes an AI-powered mobile application as well as a portable sensor that can identify and “match” green beans with a high-value taste profile.
The startup is also working with the Colombian Coffee Growers Federation to come up with a series of apps that can help farmers control and track bean quality, thereby allowing them to price their crops accordingly.
According to Eliran Lazar, VP product at Demetria, one of the goals is to enable everyone in the supply chain – from farmer to barista – to “know what they have in front of them.
“What we want to do is assist them and give them tools to make it easy,” Lazar told The Media Line.
To this end, Demetria has launched two mobile applications for Apple and Android devices. The first, called Demetria Match, is a matchmaking solution designed to help coffee suppliers properly identify coffee beans that fit their taste and quality profile. Along with the app, clients use a pocket NIR spectrometer to scan a sample of parchment or green coffee beans for average humidity levels.
The second app, Demetria Vision, uses computer vision and AI to conduct a visual analysis of beans and uncover defects. It relies on big data to reach a conclusion.
“We took a lot of samples, I think more than 2,000 photos, and then just taught the computer how it works,” Lazar explained.
After the beans are spread on a table against a dark matte background, users simply take a photo with their smartphone. The app then analyzes the image and relays an immediate result about the quality of the product.
Farmers didn’t have ways to know what kind of coffee they are really creating
Shoval, an Argentine-Israeli serial high-tech entrepreneur and angel investor, called the solution a game-changer for the industry.
“Farmers didn’t have ways to know what kind of coffee they are really creating,” he said, adding that the goal was not to replace cuppers but to bring the production and quality control process of the coffee industry into the 21st century.
“It’s good for farmers, it’s good for the cooperatives that are working with farmers, it’s good for the traders and it’s good for the roasters,” Shoval asserted.
Tens of millions of people across the world make a living from growing coffee, with the majority living in developing countries. In fact, coffee is a crucial part of several countries’ economies, including Brazil, Vietnam, Indonesia and Colombia, who lead the world in production, according to the United Nations’ Food and Agriculture Organization (FAO).
Artificial intelligence is playing an increasingly important role in the food and agricultural sectors. FAO, for instance, has adopted a number of AI and machine learning solutions to counter plant pests, detect agricultural stress and monitor and report on forests and land use.
So far, Demetria has closed $3 million in seed funding led by the Latin American-Israeli investor Celeritas and a group of private investors in the Colombian financial sector.
“It’s hard to believe that the world’s biggest roasters have effectively been buying beans with very limited knowledge about their quality, and that the majority of coffee farmers, the most critical players in the supply chain, don’t understand the quality of their own crops and hence are paid unfairly, threatening the sustainability of this $450 billion industry,” Demetria’s co-founder and CEO Felipe Ayerbe said in a statement that was shared with The Media Line.