Close Ad

The Spoon

Daily news and analysis about the food tech revolution

AI

Demetria Raises $3M to Automate Coffee Bean Analysis

Demetria, a startup that promises to automate the analysis and grading of coffee beans, came out of stealth yesterday and announced that it has raised $3 million in seed funding. The round was led by Celeritas and a group of private investors including Mercantil Colpatria, the investment arm of Grupo Colpatria.

As coffee beans move through the supply chain, their quality has traditionally been judged by “cupping.” In this process, a human with proper certifications selects samples of beans and judges them based on factors like aroma, acidity and flavor. As you can imagine, this process is slow, wasteful, and because it’s done by experts, not globally available.

It also means that coffee bean quality and pricing is a subjective process, which can incorporate any number of human biases that can affect the prices paid to farmers and across the supply chain.

Demetria aims to automate this process by using near-infrared scanning and cloud-based artificial intelligence analysis to develop “digital fingerprints” of coffee beans. As green coffee beans move through the supply chain, they are analyzed with a near-infrared scanner to look for biochemical markers to match a bean’s profile with an industry standard set of quality metrics.

This means that bean quality can be quickly assessed with a handheld scanner and mobile phone. What’s more, beans do not have to be taken out of the supply chain to tested via cupping. Instead, they stay, reducing waste.

We’ve seen this type of AI-based scanning in the food supply chain elsewhere. Most recently, Driscoll’s announced that it was using Consumer Physics’ SCiO technology to scan berries for sweetness. Consumer Physics’ handheld scanner is one of the tools being used by Demetria.

Other companies in the space include AgShift and Intello Labs, both of which use computer vision and AI to assess food quality and bring objective grading to buyers and sellers.

In its press announcement yesterday, Demetria said it has successfully completed a pilot with Carcafe, the Colombian coffee division of agricultural commodity traders Volcafe/ED&F Man. Demetria said it is also working with Federación Nacional de Cafeteros (FNC), the Colombian National Federation for Coffee Growers, to develop apps that help farmers and their transaction points in the supply chain control and track bean quality, and price.

Technologies like Demetria’s can hopefully bring more fairness to the food supply chain by speeding up the process and standardizing the analysis so everyone gets paid a fair price.

Vivino Raises $155M Series D for its Wine App and Online Marketplace

Mobile wine app and online marketplace, Vivino announced today that it has raised a $155 million Series D round of funding. This round was led by investment firm Kinnevik and Sprints Capital, with participation from GP BullHound and Creandum. This brings the company’s total funding to $221 million USD.

This most recent round of funding will be used for Vivino’s continued geographical expansion and to incorporate more personalized AI (artificial intelligence) into its platform. The company will focus its expansion efforts in markets in the US, UK, Germany, Portugal, and Japan. Its AI will be improved to create better wine recommendations for its customers.

Vivino’s community can use the mobile app or online marketplace to see reviews and ratings of different wines, and view personal recommendations based on their preferences. Customers shop for wine online through the website marketplace or app, and have wine shipped directly to their homes. Vivino allows the customer to shop for wine based on grape or region, and offers wine pairing recommendations for a variety of foods. Once the wine delivery arrives, customers can use the accompanying app to scan each bottle and receive additional information about the wine. For those unsure which bottle of wine to pick out, Vivino uses its AI-based recommendation platform to make suggestions.

E-commerce alcohol sales have significantly increased since the start of the pandemic, and this sector seems to be having its moment. Vivino’s fundraise comes just one day after alcohol delivery service and online marketplace Drizly was purchased by Uber for $1.1 billion.

Currently, 50 million people worldwide are part of the Vivino community of wine drinkers. The Vivino app is free, and available for both Apple and Android phones.

Foodspace is Using AI to Create Better CPG Data So You Find That Spicy Cheese Faster

You ever search for a food product online or at the grocery store but can’t find that exact something that perfectly matches your taste, dietary or nutritional preferences? You’re not alone. One of the big reasons searching for food products can be so frustrating is they are often bucketed under data categories that are holdovers from existing category management systems built fifteen or twenty years ago.

A new startup called Foodspace wants to eliminate this annoying experience by helping the CPG and food retail industry update their old-school category management systems with technology that makes sure that every conceivable product attribute a consumer may be searching for is documented and assigned to products headed to a physical or digital shelf.

The Boston-based startup plans on doing that by using machine vision technology that analyzes scanned images of new product packaging introduced by CPG manufacturers and uses AI to synthesize and assign attributes based on its understanding of the product packaging and label data. The attributes go beyond the typical high-level product categories such as organic or gluten free, and factors in things such as sensory preferences (creamy, grainy, etc) and consumer taste and lifestyle archetypes. All told, Foodspace’s system can assign nearly three thousand different attributes to a product.

The end result should be faster, more personalized searches for consumers. If, for example, a person who likes cheese, loves spicy food, and has a gluten allergy heads to the deli section of an online grocery store, they shouldn’t have to drill down five categories deep within the deli category to find that gluten-free habanero cheddar. With Foodspace’s AI-powered synthesis and matching of different attributes, a consumer finds a product match much faster, perhaps almost immediately, depending on the understanding the online grocer has about the shopper.

Of course, this move towards more granular, highly-consumer centric data is something that CPG and retail industries recognize is important, but have been slow to evolve away from because of the huge magnitude of switching towards systems that have thousands of product attributes. The Food Industry Association (which goes under the acronym FMI), has been working on a new framework called Shopper Centric Retailing that would update product information in the more detailed way, and this week at FMI’s annual midwinter meeting, the industry consultant who developed Shopper Centric Retailing framework, Winston Weber, announced Foodspace as a “premier” strategic solution partner to help food product companies transition their products to the new format.

In short, Weber sees Foodspace’s technology as an enabling platform to help food brands migrate to the future.

Foodspace’s technology is “helping translate products in the online space, to the benefit of brands, retailers and the end consumers,” said Weber CEO and namesake Win Weber in the press release. “Their technology is the conduit for which the Shopper-Centric Retailing business model can optimize consumer satisfaction.”

As I thought about better product data that could personalize my food product searches, I started to wonder if this could help usher in the personalized food profile concept that I’ve been thinking about ever since I heard Mike Lee talk about the idea at Smart Kitchen Summit in 2017.

Foodspace CEO Ayo Oshinaike thinks so. “The universal data set that enables that is not there,” Oshinaike told me via Zoom. “That’s the piece that’s in the middle that Foodspace is trying to solve with the breaking down of the information accuracy and how we’re able to relate products to consumers.”

Ukko Raises $40M to Fight Food Allergies and Develop its Good Gluten

Ukko, a biotech company that uses artificial intelligence (AI) to develop food and therapeutics that fight food allergies, announced today that it has raised a $40 million Series B round of funding. The round was led by Leaps by Bayer (the impact investment arm of Bayer), with participation from Continental Grain Company, PeakBridge Ventures, Skyviews Life Science and Fall Line Capital. Existing investors including Khosla Ventures and TIME Ventures, the investment fund of Marc Benioff, participated as well.

This brings the total amount of funding raised by Ukko to $47.7 million dollars. According to the press announcement, the new funding will allow Tel Aviv, Israel-based Ukko to enter into clinical trials for its investigational therapeutic for peanut allergies and, more relevant for our purposes, accelerate development of the company’s proprietary gluten.

Simply speaking, Ukko uses its AI platform to analyze patient data to map how an allergen triggers a reaction in the body. With that information, Ukko breaks down the gluten protein to its component level and gets rids of the bad parts that cause allergic reactions. It keeps the good parts. Ukko then creates this new good gluten either by genetically modifying wheat plants or fermenting yeast (or some other applicable base cell) to grow it in a bioreactor.

The result, Ukko Co-Founder and CEO Anat Binur told me by phone this week, is a gluten that stretches and bakes and has all the biophysical aspects of gluten, and can be eaten by people with gluten sensitivities and celiac disease. This, in turn, means that gluten-sensitive people don’t need to sacrifice quality when enjoying different types of baked goods.

At least, that’s the plan. Binur said that some of the company’s new funds will go towards clinical trials of its gluten and getting the product through all the safety protocols and to the point of commercialization.

Once Ukko’s gluten reaches the commercialization stage, Binur said that there are a number of options for how it comes to market. Ukko could sell its own gluten, which could be added to gluten-free starches (like almonds or rice). Alternatively, the company could sell its own gluten flour as an ingredient to food companies and restaurants/bakeries, or create its own line of branded gluten-free flour to be sold on store shelves. Or Ukko could pursue some combination of all three.

By one estimate from Grand View Research, the gluten-free products market was valued at $21.61 billion in 2019, and projected to grow at a CAGR of 9.2 percent through 2027. So there is plenty of market opportunity just in gluten for Ukko.

But Ukko’s platform can be applied to any food allergy. As noted, the company is developing therapuetics for peanut allergies, but Ukko’s tech could be used for dairy allergies, soy allergies, egg allergies, etc. Creating replacement foods from the ground up that have the same nutrition and behave like the original could help alleviate a lot of sickness and save lives.

Google Cloud Team Uses AI to Develop Hybrid “Cakie” and “Breakie” Recipes

I remember at some point last year after the pandemic had started, I called three different grocery stores to check if they had yeast in stock. Of course, I had no luck at any location because our country had entered a baking extravaganza; ovens were fired up non-stop across the country to bake sourdough, banana bread, and chocolate chip cookies. Due to this baking frenzy that overtook our country in 2020, the Google Cloud team analyzed the ingredients and specific ratios used in favorite baked goods like cookies, bread, and cakes through the use of a machine learning program. The goal was to essentially determine what defined something as a cookie, cake, bread, and how each of these categories differ from each other. As a result, the team also used its AI to produced two new hybrid baked goods called a “Cakie” and a “Breakie”.

The team collected hundreds of different recipes for bread, cookies, and cakes to create a dataset and then used Google Cloud’s AutoML Tables tool to build a machine learning model. The model analyzed the amounts of key ingredients such as yeast, butter, eggs, and sugar, in the different recipes and was able to determine if the recipe was for bread, cake, or cookies. Bread recipes were accurately labeled about 90 percent of the time, but sometimes the model mislabeled cookie recipes for cake recipes.

Using this model, the team used ingredients and measurements that the algorithm determined were key for cookie, bread, and cake recipes to create two new hybrid recipes. The “Breakie” is half cookie and half bread, which turned out to be a fluffy, airy cookie with a texture similar to that of a muffin. The “Cakie” maintained a cake shape and sponginess but has a buttery flavor and crispy outside like a cookie.

Google Cloud’s project is just one neat example of how machine learning and AI (artificial intelligence) can be applied in food technology. Spoonshot recently launched a free version of its AI flavor-pairing tool that offers the user both novel and unique pairings for different ingredients. Brightseed created an AI platform called Forager, which detects hidden phytonutrients in different plants. Even companies like Starbucks, Sweetgreen, KFC, and McDonald’s use AI to some degree in order to streamline day to day operations.

As we still find ourselves in the pandemic early 2021, you may be looking for some new recipes to bake. You’re in luck; the Google Cloud team provided the two recipes created through the use of AI, the “Cakie” and the “Breakie”.

Spoonshot Launches Free Version of its AI-Based Flavor Pairing Tool

Spoonshot, which uses artificial intelligence (AI) to uncover novel flavor combinations, has launched a free version of its tool that is accessible to anyone. Spoonshot CEO and Co-Founder, Kishan Vasani spoke about the new level of service at The Spoon’s Food Tech Live event earlier this week.

Up to this point, Spoonshot’s platform has been a B2B play, meant for CPG companies and foodservice operators looking ahead to see what the next food and flavor trends might be. As we wrote last year when Spoonshot raised $11 million:

To get ahead of the curve, Spoonshot’s platform examines data from across a vast number of food-related sources including online menus, food science, CPG ingredients and online food communities. Spoonshot runs this data through its proprietary machine learning and AI algorithms to help companies identify existing and novel flavor combinations.

The key word here is “novel.” When you enter a flavor like “banana” into Spoonshot’s Ingredient Network tool, the service brings back a number of potential flavor combinations and scores them based on novelty. Combining banana + chocolate is common, but combining banana + aloveera juice is something that probably hasn’t occurred to most people and a combo that Spoonshot says will be tasty.

Armed with this novel combination, a restaurant or CPG company could go about building a new product that will appeal to consumers.

Launched at the start of the new year, Spoonshot’s new free tier of service now allows anyone to try its AI platform out (pricing starts at $99/month). Chances are good that most of us in our everyday lives don’t need enterprise-grade artificial intelligence to uncover novel flavor combinations. But aside from being a fun (and free) little distraction for everyday chefs, it could also be useful for small CPG or restaurant owners that don’t have R&D budgets to expand their offerings.

Brightseed’s First Major Phytonutrient Discovery Finds Black Pepper May Help with Fatty Liver

Brightseed, which uses artificial intelligence (AI) to uncover previously hidden phytonutrients in plants, today announced preclinical data from its first major discovery targeting liver and metabolic health.

The discovery was made with Forager, Brightseed’s AI platform that looks at plants on a molecular level to identify novel phytonutrient compounds (for example, antioxidants in blueberries). Once found, Forager then catalogs these compounds and uses that information to predict the health benefits of those compounds.

With today’s announcement, Brightseed’s Forager has identified phytonutrients that can help with fat accumulation in the pancreas and liver, a condition linked to obesity. Brightseed explained its findings in a press release, writing:

Using a computational approach with data from Brightseed’s plant compound library, Forager identified two natural compounds with promising bioactive function, N-trans caffeoyltyramine (NTC) and N-trans-feruloyltyramine (NTF). Researchers determined that these compounds acted through a novel biological mechanism governing the accumulation and clearance of liver fat. The preclinical data was presented in the fall of 2020 as a poster session at The Liver Meeting® Digital Experience hosted by American Association for the Study of Liver Diseases, and published as abstract #1679 in Hepatology: Vol 72, No S1. 

The release continued:

IIn preclinical studies, NTC and NTF acted as potent HNF4a activators, promoting fat clearance from the steatotic livers of mice fed a high fat diet, by inducing lipophagy.  HNF4a is a central metabolic regulator that is impaired by elevated levels of fat in the bloodstream resulting from chronic overeating. Administered in proper doses, NTC and NTF restored proper function of this central metabolic regulator, including maintaining healthy lipid and sugar levels in the bloodstream to normalize organ function. Their activities were confirmed using a cell-based human insulin promoter activation assay. Forager found NTC and NTF in over 80 common edible plant sources. 

One of those plant sources, Brightseed Co-Founder and CEO, Jim Flatt told me by phone this week, is black pepper. Now, before you run out and grab your pepper grinder, there is still a lot of work that remains before the results of this discovery bear out.

First, the compounds still need to go through clinical trials to validate Brightseed’s initial findings. This includes not only confirming any health benefits, but also determining the doses and best methods for administering the compounds. Then the best plant source for those compounds needs to be determined as well as the best method for compound extraction. Flatt told me that if all goes well, you can expect to see some form of supplement on the market by the end of 2022.

Even though that is a ways off, part of the reason to be excited by today’s announcement is because of how little time it took Brightseed to make this particular discovery. Through its computational processes, Flatt told me his company was able to shrink what used to take years down to months. “Fifteen to 20 percent of time that is computational saves us 80 percent of the time in the lab,” Flatt said.

Brightseed has already analyzed roughly 700,000 compounds in the plant world for health properties and says it’s on track to surpass 10 million by 2025. Doing so could help unlock a number of previously unknown treatments for a number of ailments and conditions as well as general improvement to our metabolic and immuno health.

In addition to independent research such as today’s findings, Brightseed also partners with major CPG brands to help them identify new applications for their products. For instance, Danone is using Brightseed’s technology to help find new health benefits of soy.

Brightseed’s announcement today also reinforces the bigger role AI will play in our food system. AI and machine learning is being used to do everything from turning data into cheese, to solving complex issues around protein folding.

As more discoveries using AI are made, more investment will be poured into the space, which will accelerate even more discoveries.

Sony AI Unveils Trio of Food Projects Including AI-Powered Recipes and Robots

Sony AI, an artificial intelligence and robotics research and development company spun out from Sony, today announced its “Gastronomy Flagship Project.” The new food-related endeavors include an AI-powered recipe creation app, a robot chef’s assistant and a community co-creation initiative.

Sony launched Sony AI last November and the unit became its own company in April of this year. Sony AI’s mission is to develop AI-related products for videogames, imaging and sensing, and gastronomy “with the aim of enhancing the creativity and techniques of chefs around the world,” according to today’s press announcement.

To that end, the three projects Sony AI announced today are:

  • AI-Powered Recipe Creation App – Gathering up data such as aroma, flavor, molecular structure, nutrients and more, this app will use AI algorithms to help chefs create novel food pairing, recipes and menus.
  • Chef Assisting Cooking Robot – This is pretty self-explanatory. Sony AI will develop a robot that can mimic the physical actions of a human chef to do everything from preparing to plating dishes.
  • Community Co-creation Initiative – This project is a little more vague, with Sony AI only saying it will “aim to contribute to the long-term sustainability of the community” through relationships with universities, research institutes, companies and more. The first step in this process is a “Chef Interview” video series on the Sony AI website.

Sony AI’s press release didn’t provide specific details around when or where these projects would launch (other than the video series).

But today’s announcement continues a trend we’ve been seeing with large electronics corporations doing advanced work on food robotics and AI. LG is working on robot waiters with Woowa Brothers in South Korea. And Panasonic is working with with the Haidilao chain of hot pot restaurants in China to develop a robotic kitchen.

Sony itself is no slouch when it comes to food-related robots. The company collaborated on cooking robot research with Carnegie-Mellon University a couple years back. And it has a pretty grand vision for advanced AI-powered cooking robots and assistants that could make Michelin star meals in your home.

How Alphabet’s AI-Powered Leap in Protein Structure Prediction Could Accelerate New Food Development

This past week Alphabet, the parent company of Google, announced that its DeepMind group has solved a long-standing grand challenge in the scientific community around protein structure prediction.

In 1972, scientist Christian Anfinsen predicted that a protein’s structure could be determined by its amino acid sequence. However, figuring out that sequence is immensely difficult since there are near-infinite ways in which a protein can fold. This led Cyrus Levinthal to postulate that calculating all the known configurations would take longer than the age of the known universe if our only way there was brute force calculation (a problem often referred to as Levinthal’s Paradox).

Thankfully, now we won’t have to wait forever (literally) since DeepMind’s DeepFold AI can predict protein structure to the width of an atom within days. My former Gigaom colleague, Darrell Etherington, explained the feat in a post on Techcrunch:

The test that AlphaFold passed essentially shows that the AI can correctly figure out, to a very high degree of accuracy (accurate to within the width of an atom, in fact), the structure of proteins in just days — a very complex task that is crucial to figuring out how diseases can be best treated, as well as solving other big problems like working out how best to break down ecologically dangerous material like toxic waste.

As many in the world of future food development know, AI is becoming an increasingly important tool to accelerate the development of new proteins and other food-building blocks. Companies like Climax Foods are embracing machine learning to help them develop new approaches to making food like plant-based cheese. And now, with DeepMind’s advances in protein folding prediction paths, we can expect AI to become an even more important tool in new food development.

To get a feel for what the impact of this milestone might be, I asked Sudeep Agarwala, a synthetic microbiologist for Gingko, what he thought DeepMind’s work could possibly mean for food:

“There’s so much that can happen with this,” Agarwala told me. “Think about different textures or mouth feels with food proteins that the AI can design. Or even different amino acid contents for the proteins. And that’s just for the end proteins (if you want the protein as the end product).”

“If we’re engineering proteins inside the cell to produce a small molecule (a flavor, a fragrance, a fat) we can think about making completely new enzymes with multiple functionalities. So something like a 5-step process might be condensed. This has the potential to simplify so much of the metabolic engineering we do to produce these products.”

Agarwala also believes that using technology like that of DeepMind’s will also be way more effecient in terms of resources and have a much smaller ecological footprint:

“Even if the proteins aren’t ultimately going to be consumed, being able to rationally design enzymes that are involved in producing small molecules (think fats, fragrances, or flavors) is really exciting to think about,” said Agarwala. “This isn’t small potatoes: think about how much energy and resources goes into making vanilla or saffron, for example. Being able to do this more efficiently will provide a less ecologically expensive way to produce these materials.”

Just as disease and virology experts are buzzing about what this new milestone could mean for their work, it’s clear food scientists like Agarwala are equally as excited.

“It’s such an exciting time to be a biologist!” said Agarwala.

Yes, it is.

Kea Raises $10M to Bring its AI-Based Phone System to More Restaurants

Kea, which makes an AI-powered virtual phone assistant for restaurants, has raised a $10 million Series A round of funding. TechCrunch was first to report on the round, which was led by Marbruck, with Streamlined Ventures, Xfund, Heartland Ventures, DEEPCORE, Barrel Ventures and AVG Funds, as well as other angel investors participating. This brings the total amount raised by Kea to $17 million.

Basically, Kea is building an automated way for restaurants to answer the phone. The natural language processing software can hold a “conversation” with a customer to take and process their order. You can hear a demo of how it works on the Kea website.

As TechCrunch points out, many restaurants are understaffed and don’t always have a dedicated person to work the phones and take orders. During this time of off-premises eating, not answering the phone can translate into a lot of lost business. Plus, ordering directly from the restaurant instead of through a third-party like DoorDash or Uber Eats helps the restaurant avoid the sky-high commissions those services charge.

Kea is among a wave of natural language customer interaction systems coming to market. Google famously made news a couple years back with its Duplex AI-powered voice assistant for consumers to make automated restaurant reservations that sounded almost too human. Google also developed the CallJoy virtual phone assistant for small business owners. Clinc’s technology brings natural language conversations to the drive-thru, while McDonald’s acquired Apprente last year to add more voice capabilities to its drive-thru.

Kea told TechCrunch that its service is currently live in more than 250 restaurants including Papa John’s. With its new cash, the company is looking to be in 1,000 restaurants in 37 states next year.

SKS 2020: Grabango Says its Computer Vision Avoids Biases Because it Doesn’t Use Facial Recognition

Grabango CEO Will Glaser said this week that because his company’s technology doesn’t use facial recognition, it can avoid some of the same types of racial biases found in other computer vision and machine learning algorithms. Glaser’s comments came during my fireside chat with him at our Smart Kitchen Summit this week.

Grabango makes cashierless (or as Grabango calls it, “lineless”) checkout technology for grocery retail. Grabango installs hundreds of cameras in the ceiling of a store, and through a combination of computer vision and AI, keeps track of what people put in their basket (or pocket). When shoppers go to check out, they choose a cashier or use an accompanying mobile app that automatically charges for the items kept.

This type of computer vision-based system is also used by other cashierless checkout companies such as Zippin, Standard Cognition and Amazon, and could become more widespread as demand for contactless retail experiences grows.

But the problem with some computer vision + AI systems is that their algorithms can contain the human biases of their creators. As TechCrunch wrote earlier this year, “MIT researchers reported in January 2019 that facial recognition software is less accurate in identifying humans with darker pigmentation.” In a busy retail environment where a lot of people are grabbing a lot of different items all throughout the day, it’s not hard to see how this type of bias could adversely impact people of color.

Glaser recognized some of the problems that can arise when algorithms are only developed or trained on white men in a lab. He made a point of saying that Grabango’s technology does not use facial recognition, and that it has a 99.6 percent revenue accuracy rate. Grabango’s system does follow a person around store, but doesn’t personally identify them (it can, however, stay with a person even after they have put on or taken off a jacket).

Grabango’s system is also being used in real world conditions, so it continues to train its algorithms in a diverse setting. Last month, Grabango announced that it’s integration with Giant Eagle’s GetGo Market in the Pittsburgh area went live. So now we’ll be able to see in a more open environment if Glaser’s claims hold true.

The Food Tech Show: Did The Automat Ever Really Go Away?

In this week’s episode of The Food Tech Show, we talk about those new contactless systems and compare them to a technology from long ago: the automat.

Yep, that old-school idea born in New York City a century ago is back (or maybe it never left?), showing up everywhere from restaurants to condos.

Jenn Marston waxes nostalgic about the automat and other concepts that seem to be getting a second look as the food system looks to reinvent itself in the wake of COVID-19.

We also talk about these stories in today’s podcast:

As always, you can listen to this week’s podcast on Apple Podcasts, Spotify or wherever you get podcasts. You can also download the episode direct to your device or just click play below.