Strella Believes Its Learning Machine Will Help Deliver Perfectly Ripe Bananas
Did you know there is a job in the banana industry called a ripener?
Makes sense, right? After all, anyone who eats bananas knows the time it takes to go from a green banana to a brown waste can be as short as a week. This means that the banana industry has to work hard to ensure that the bananas ripen at the right time so that they appear bright, beautiful yellow when they appear on the grocery store shelves.
Like most jobs, the role of maturity relies heavily on judgment. It’s not that they don’t use other modern tools when monitoring and controlling the banana ripening cycle, but from the looks of it, the ripening task seems ripe (sorry) with a Moneyball-style analytical and technological revolution.
Enter the Strella. The company, which has gained traction in the apple industry with its IoT monitoring technology over the past few years, has gone bananas. According to the company’s CEO Katherine Sizov, the company’s new AI-powered model helps them (and those who work as adults) better understand the signals that bananas send.
“We built a machine learning model that helps us detect bananas from that green to the perfect yellow every time,” Sizov told Spoon. “And the way we do it is we measure what the bananas tell us.”
According to Sizov, the hardware they use to monitor bananas is similar to apples. The difference is the software.
“The hardware is the same, but the algorithms are different,” Sizov said.
Sizov says that whether it’s apples or pears (fruits with a long ripening cycle) or avocados or bananas (fruit with short ripening cycles), the important indicators that send signals of the ripening stage are ethylene and CO2 emitted from the product. The Strella hardware module has eight different sensors, sensing ethylene, CO2, and other environmental factors such as temperature and humidity.
And like apples, Strella technology can help determine what exactly is needed to slow down or speed up the ripening cycle of bananas. The only difference is that things move much faster with bananas or avocados, which is why the work is clearly focused on controlling the ripening process of the product.
“Unlike bananas, apples grow well on the tree,” said Sizov. And they can last year round in large storage sheds.”
With bananas or avocados, the ripening process is very closely controlled. They are ripened before they are ripe and kept cold to reduce the ripening until they reach the point of use. From there, they enter the ripening rooms, where the ripener injects ethylene gas and CO2 and adjusts the temperature to kick the ripening process into gear. And now, according to Strizov, the new Strella banana and avocado machine learning algorithm can help determine precisely how much each is needed to adjust the ripening cycle to get the desired output.
Should adults be concerned about technology taking over their jobs? Sizov does not think so.
“When people excel at their work, they always look for tools to do better,” said Sizov. “Adolescents have a ton on their plate, working 12- to 14-hour shifts, so I think they’re always looking for ways to get more sleep. Our tool is one way to do that.”
According to Sizov, Strella has worked with 85% of the US market for apple and pear suppliers and estimates that the company has saved 20 million pounds of apples and pears from waste. Now, he hopes they can replicate that success in bananas and avocados.
“We’re growing very quickly, and we’re excited to get into bananas and avocados after taking apples for five years now.”
If you’d like to hear Katherine discuss how AI can perfect banana ripening, she’ll be speaking at the Spoon’s Food AI Conference on October 25th in Alameda, CA! Get your early bird tickets today!
