Altair Engineering Inc.

10/29/2024 | News release | Distributed by Public on 10/29/2024 05:04

Future Says S5E14 Recap: Fighting Food Waste with Machine Learning and AI

Food waste, though sometimes seen as an individual problem, is one of society's great systemic issues. According to the nonprofit hunger relief organization Feeding America, 92 billion - that's 92,000,000,000 - pounds of food go to waste each year in the U.S alone. In other terms, 38% of all food in the U.S. goes unsold or uneaten, causing 145 billion meals' worth of food to end up in landfills. Moreover, in addition to being a social and economic problem, food waste is also a climate problem, representing nearly 8% of all global carbon emissions. And beyond the facts and figures, food waste represents one of modern society's great contradictions. Despite this abundance of food, according to the World Health Organization, as many as 828 million people worldwide were affected by hunger as recently as 2021. In short, though our world produces more than enough food for all, hundreds of millions of people face hunger daily.

Remedying the multifaceted challenge of global food waste is what drove our latest guest on Future Says, Katherine Sizov, to start her own company: Strella Biotech. In fact, Feeding America's data - specifically the 38% statistic - was the catalyst from which Strella emerged. "I thought it was a number that doesn't make sense in today's world, so I set out to discover why waste occurs and build a solution," Sizov says.

Tackling Perishability, Intercepting Hidden Signals

Food waste occurs because of a variety of factors. One of these is perishability, referring to the process through which food spoils. Perishability is a key concern wherever fresh produce is concerned. In the simplest terms, Strella wants to minimize unnecessary spoilage. According to Sizov, "What we do is predict the shelf life of fruits and vegetables and use that data to optimize inventory across the supply chain. We do this by using a combination of machine learning and sensor technology that measures different gases that fruits and vegetables produce. Based on these gas profiles and emissions, we can predict where that fruit or vegetable is [in terms of maturation]. From there, we can build an algorithm that can tell us what the expiration date looks like." In other words, Strella uses sensors, machine learning, and artificial intelligence (AI) to optimize how fruits and vegetables are stored, presented, and sold.

Strella's sensors are about the size of a brick and can be placed wherever fresh produce is located, whether that be on a shelf, in a box or vehicle, inside a container, and beyond. Once placed, their job is essentially to "listen" to the produce they're monitoring. "Fruits and vegetables 'talk' to one another all the time," Sizov says. "For example, you can see this when you put an unripened banana next to a ripened one, which causes the unripened banana to ripen faster than normal. Basically, these sensors measure the 'secret communications' between fruits and vegetables. Our sensors intercept these signals and tell us what the foods are telling each other."

Addressing a Crucial Niche

To combat food waste, Sizov founded Strella to address two glaring issues she sees in the modern produce supply chain: not having enough data and the prevalence of silos. As she sees it, not enough data is being gathered throughout the produce life cycle, which makes estimating freshness an imprecise guessing game. And secondly, she says that though many large retailers sell fresh produce, they don't necessarily specialize in produce. In other words, they and their suppliers don't have the time, expertise, or willingness needed to optimize the way we think about produce supply chains. Sizov says this leads retailers and suppliers to essentially play a game of "hot potato" with produce - working largely independent of one another, sharing little data - until it reaches shelves, which perpetuates the creation of food waste.

Strella's goal is to be the missing link throughout the produce supply chain - a hub of food data and expertise. "What we want to do is be their data provider for all things food," Sizov says. "I think all of us as consumers can understand that the produce section is one of the most sensitive parts of the store. A lot of people pick their grocery store because of [the produce], so nailing down that category in retail is key."

In doing so, data analytics and AI have a pivotal role to play. Since all fruits and vegetables have unique characteristics - in that they produce different gases, have varying tolerance to bruising or mold, etc. - Strella creates unique machine learning models for every type of produce. That means Strella has "banana" models, "cauliflower models, "kiwi" models and more. Machine learning and AI helps the Strella team both analyze and predict how foods will behave in given circumstances, helping suppliers and retailers determine the best ways to store and sell produce. That means with the right data, Strella technology can help suppliers keep food fresher for longer, help retailers put the best possible food on their shelves for the longest amount of time, and ensure consumers are getting food that won't spoil right after they buy it.

A Universal Interest

Though Sizov says starting a company is a journey filled with challenges and learning experiences - "I've made about every mistake possible" - it's the feeling of making a difference that drives her. More than anything, she says she enjoys working in an industry that feels "real," one that has an impact on all our lives. "It's food - it's something everybody interacts with and has an interest in [making better]," she says.

As Strella grows and evolves, Sizov hopes the company is known for helping address food waste and perishability by using a combination of next-generation software and accessible hardware. And though she acknowledges that perishability is only one component of food waste, every bit of innovation and progress counts. She thinks AI will be a key tool in our toolbelt moving forward. "I think [AI] is part of the solution, but as with all big problems, you have to attack it from a lot of different angles and there's no real silver bullet," she says. "But I think AI, having good datasets, and doing good data science is crucial."

Few know the challenges of stemming the tide of food waste like Sizov and Strella. It's a monumental, global challenge. But both the company and its founder believe that a combination of passion, grit, and technology can point the path to a brighter, more sustainable, less wasteful future. "[Creating and running Strella] is definitely the hardest thing I've ever done in my life," Sizov says. "But I think part of living a fulfilling life is doing the challenging things. If I didn't like what I was doing, I would've burned out a long time ago."

Click here to listen to the full episode with Strella Biotech's Katherine Sizov. To check out the rest of Future Says season five, visit https://altair.com/future-says. And be sure to subscribe to Future Says on Spotify, Apple Podcasts, Amazon Music/Audible, YouTube Music, and Podcast Addict.