• Skip to primary navigation
  • Skip to main content
  • Skip to primary sidebar
  • Skip to footer
  • Skip to navigation
Close Ad

The Spoon

Daily news and analysis about the food tech revolution

  • Home
  • Podcasts
  • Events
  • Newsletter
  • Connect
    • Custom Events
    • Slack
    • RSS
    • Send us a Tip
  • Advertise
  • Consulting
  • About
The Spoon
  • Home
  • Podcasts
  • Newsletter
  • Events
  • Advertise
  • About

AI

September 8, 2021

Food Tech Patent Watch: Patent Reveals Eatsa’s Robotic Meal-Assembly Machine

Remember Eatsa?

You know, the automat-like bowl food restaurant that was re-spun as a fast-growing (but more boring) restaurant marketing tech company called Brightloom?

I do, mainly because I loved the place. After I visited one in New York City, I wrote that the restaurant could be the future of fast-casual dining.

As it turns out, some – including maybe Eatsa’s investors – didn’t agree with me. I say that because starting in 2019, they phased out the cubbies, changed their name, took money from Starbucks, and, from the looks of it, dropped big plans for automating the back of house with meal assembly robots.

I say that because Eatsa (now Brightloom) holding company Keenwawa, Inc. was issued a patent last month for a meal-making robot. The patent, a continuation patent for one first issued to the company in 2019, shows a system that assembles meals by dispensing different ingredients stored in canisters into bowls and then shuttles the assembled meals off to the cubbies in the front of house.

Drawing of The Eatsa Meal Assembly Machine From Patent Filing

From the patent:

The automatic food preparation and serving apparatus may also comprise the food dispensing mechanism configured to dispense the ingredient from the plurality of food canisters into the bowl or food receptacle, under program control of the one or more processors.

The patent goes into excruciating detail about the system, complete with dozens of images outlining the canisters, the dispensing system, the conveyor belt, the bowls, and even the touchscreen user interface (which looks a lot like those deployed in the actual Eatsa restaurants for consumers to choose their bowls).

The list of inventors on the 60-page plus patent includes the former automation and engineering team for Eatsa, as well as Dave Friedberg, the one-time Climate Corp founder who incubated the company as part of what would eventually become The Production Board holding company.

I have to wonder if Eatsa-now-Brightloom’s owners are looking to license or sell the technology or even revive their ‘eatsa-inside’ strategy. After all, the recent news that Sweetgreen had acquired Spyce to help do exactly what Eatsa did – make food bowls – shows that some restaurants see a future in automated meal assembly.

Time will tell. For now, however, you can take a look back at the Eatsa ordering and cubby system below.

View this post on Instagram

A post shared by Michael Wolf (@michaelawolf)

Food Tech Patent News Roundup

The First Patent Awarded to an AI System is in Food

While I thought that artificial intelligence systems will do most everything at some point, I assumed an AI being awarded a patent was one we could sleep on for a few years more.

I was wrong. In the July edition of the Patent Journal, an AI system named DABUS was awarded a patent for an innovation called “food container based on fractal geometry” for a system with interlocking food containers.

This Quartz Africa article describes DABUS’s creator:

DABUS (which stands for “device for the autonomous bootstrapping of unified sentience”) is an AI system created by Stephen Thaler, a pioneer in the field of AI and programming. The system simulates human brainstorming and creates new inventions. DABUS is a particular type of AI, often referred to as “creativity machines” because they are capable of independent and complex functioning. This differs from everyday AI like Siri, the “voice” of Apple’s iPhones.

DABUS’s inventor Thaler submitted the patent application listing DABUS as the inventor since the AI conceived and created the food storage system entirely. He submitted it to patent offices worldwide, including the US, which rejected him because, among other reasons, the US patent office only awards patents (for the time being) to human inventors. However, the South Africa patent office apparently had no such restrictions when it surprised many with an award of a patent.

In retrospect, it shouldn’t be that surprising AI are creating patentable ideas. AIs already write novels and movie scripts, so why not invent novel things like food containers and get them patented?

I’m looking forward to when we see AI start innovating on novel food. We’ve seen what an impact AI had in vaccine development, so it’s not too much of a stretch to see how it could start making a difference on the bioengineering front for foods.

A Patent Awarded for Many Container Within Container Scenarios For Food Storage, Cooking, and More

A fairly wide-ranging patent titled “Multi-function compact appliance and methods for a food or item in a container with a container storage technology” (US011104502) has been awarded to an Edward Espinosa from Spain for a system that enables a variety of container within container use-cases.

One example is a food container within a larger food storage appliance (i.e., a fridge) sending information on status such as freshness, etc. Another is the refrigerator with a microwave oven in one of the compartments.

The patent describes various technologies such as NFC, Bluetooth, voice control, machine vision via an internal camera, and more to enable the container systems to communicate with the appliance. In addition, the system describes the use of smart tags that communicate freshness data from within storage drawers.

I’ve long called for innovations in the core design of the fridge since they’ve largely been the same for the past 100 years, and it looks like Espinosa has definitely given the refrigerator a rethink.

Whirlpool’s Solid State Cooking Patent

Schematic for Cooking System With Directed RF Energy

It’s been a loosely held secret in the appliance industry that Whirlpool has been tinkering around in the solid-state cooking area for a while. This patent describing a system with multi-regional cooking via RF signals shows they are also trying to accumulate IP in the market. At this time, only Miele has commercialized a consumer solid-state cooking appliance, but hopefully, soon, we’ll see a next-generation microwave from a mass-market brand like Whirlpool.

August 23, 2021

Forget Watching TV. Sony Wants Your TV To Watch You (and Monitor Your Eating)

The couch in front of our TV is a sacred no-judgment zone, the place where we wear sweats, eat pizza, and binge-watch Cobra Kai on Netflix.

But what if the TV watched and, well, judged us? That’s the dystopian future Sony is thinking about, at least if US patent number 11,081,227 is any indication.

Called “Monitoring and reporting the health condition of a television user,” the patent describes a television with embedded cameras and sensors that monitor everything a person while slouched on the couch, including tracking the food they eat. The patent was awarded to Sony on August 3rd of this year.

So why is Sony looking to invade the safe space on our couch? According to the patent, the thinking is a TV like this could monitor the health of the TV viewer by recording activities such and eating habits and look for any signs of a potential impending health problem.

Illustration of Remote with Biometric Sensors from Sony Patent #11,081,227

From the patent:

People spend a significant number of hours sitting in front of a television (TV). They may have a hidden health problem or a diminishing health condition. The system monitors one or more health vitals of the user such as heart rate, etc. While sitting in front of the TV, they may behave in an unhealthy manner. For example, they may eat too much while watching a TV program. The system may also monitor the types of food a person eats while watching TV.

The patent describes a system that would use a camera embedded in the TV to log a person’s food consumption and monitor eating patterns such as chewing speed, posture, and how fast they eat. The system would also identify the type of food, estimate portion size, and take a guess at how many calories are consumed. All that info could be synced with biometric monitoring (the patent suggests a heartbeat sensor in the remote) to paint a profile of a person’s health and how their TV watching habits impact it.

So who needs a TV to monitor their food intake? I guess if one consumes lots of empty calories while sitting in front of the tube, this might prod them to change their habits. But on the other hand, do we really need a TV to tell us we’re eating poorly?

And then there’s the problem of consumer privacy. Some of you may remember the controversy a few years back about a few lines in the terms of service or a Samsung smart TV that suggested the company might capture conversations and sell them to third parties. I can only imagine what people might think about cameras on their TV watching their every move.

In Sony’s defense, this isn’t close to being an actual product. Patents are oftentimes just corporate thought experiments, where R&D managers with budgets to burn ask what if? The likelihood of Sony actually building their big brother TV is probably pretty low.

Probably. But maybe someday they will, and if they do, it’ll be interesting to see if consumers are ok with the idea of a TV watching them, trying to make them better people. But hey, if you’re really watching Cobra Kai and downing whole quarts of ice cream every Friday night, maybe you could use the help of your Sony 70″ 4K OLED TV to help perform a little self-care.

August 12, 2021

Sama Provides the Data to Fight Food Waste and Power Cashierless Checkout

When we talk about artificial intelligence (AI) in food tech, it’s often about the end result: Cashierless checkout, crop assessment, autonomous vehicles, etc. But one thing that these solutions and any other using AI need is is good data. Sama is a company in the good data business, and it has built a platform that provides training data that other companies can use to speed up the development of their AI models.

I spoke with Wendy Gonzalez, CEO of Sama, this week by video chat. She outlined some of the food tech use cases for her company’s technology, such as fighting food waste. “If you’re in a restaurant or hotel, catering service. A lot of that food gets wasted,” Gonzalez said. Sama is working with a company called Orbisk that provides a device for commercial kitchens that uses computer vision to analyze the food being thrown away. For example, if Orbisk sees a lot of mac and cheese is being tossed because no one is taking it, that kitchen can know not to make as much of it, and by extension save money by ordering less of the ingredients to make mac and cheese. (Winnow is another company that takes this same approach.)

For its part, Orbisk had a thousands of images of different types of food for its system to recognize. Sama came in and provided structure and taxonomy to that data. In other words, Sama labeled all the images of mac and cheese accordingly to train Orbisk’s AI to automatically recognize mac and cheese. The result, according to Sama, is that Orbisk’s system can reduce food waste in commercial kitchens by as much as 70 percent.

Gonzalez said that Sama’s system is also being used in other fields like cashierless checkout. In that setting, Sama is helping train those computer vision systems to recognize packaged goods, which is more complicated than people think. A cashierless checkout system needs to not only recognize a package of Oreos, but a package of Oreos in different lighting conditions, different angles or when the view is partially blocked. Sama provides all of that data.

Sama is also being used to train AI systems on early crop disease detection, automated crop harvesting, and soil condition monitoring.

Sama is among a number of players in the data space including AI.Reverie, which uses synthetic data to create images virtually to train AI models, and Nvidia, which is also using synthetic data to train robots to navigate around a kitchen.

As AI plays a bigger role in our everyday lives, there will be a growing need for more good data to train those AI systems. And the end result is that we’ll all be talking about good data more often.

August 4, 2021

Danone Expands Brightseed Partnership to Uncover Hidden Healthy Compounds in Plants

Danone and Brightseed announced an expanded partnership today that will have the food and beverage giant using Brightseed’s Forager artificial intelligence platform to uncover more phytonutrients from additional plant-based ingredients.

Brightseed’s technology studies plants on a molecular level to identify and catalog previously unknown compounds that could have health benefits. For example, earlier this year Brightseed announced that it had discovered in pre-clinical trials that the bioactive compounds N-trans caffeoyltyramine (NTC) and N-trans-feruloyltyramine (NTF) found in black pepper can help with the clearance of fat accumulated in the liver. After these initial findings through Brightseed’s AI, the company will move forwards to confirm the results through clinical trials to determine efficacy as well as other factors such as dosage and administering the compounds.

Danone first teamed up with Brightseed in June of last year to study potential new benefits of soy. (Danone owns the Silk brand.) According to today’s press announcement sent to The Spoon, Brightseed’s Forager discovered 10 times more bioactives than previously known and 7 new health areas. As these findings are confirmed by more clinical data, brands like Danone benefit because they can tout additional health benefits around their products, but consumers benefit because there are then more plant-based tools to fight different ailments.

What’s interesting about companies in the AI space like Brightseed, Spoonshot, and Journey Foods is how they are shortening the discovery period for food companies looking to create new products. Before machine learning and artificial intelligence, food manufacturers had to first hypothesize about how particular ingredients might work together, or the health benefits of an ingredient. After that guess, they would run physical tests in a lab to see if they were remotely close in their hypothesis. If they were wrong, they’d have to start all over again from scratch. AI helps shrink that time by doing a lot of that guesswork up front quickly in a computer before any lab time is needed.

The whole space is very new, but Danone expanding on its partnership with Brightseed is a vote of confidence for the technology and should lead to more brands jumping into the use of AI and computational biology.

July 29, 2021

ConverseNow Raises $15M for Restaurant AI Assistant Tech

ConverseNow, which makes an AI ordering assistant for restaurants, announced today that it has rasied a $15 million Series A round of funding. The round was led by Craft Ventures with participation from existing investors including LiveOak Venture Partners, Tensility Venture Partners, Knoll Ventures, Bala Investments, 2048 Ventures and Bridge Investments. This brings the total amount of funding raised by ConverseNow to $18.3 million.

ConverseNow’s lead product is its voice-based assistants dubbed George and Becky. These automated assistants can understand and take orders from customers who simply say what they want as they pull up to a restaurant’s drive-thru. During a recent video chat, ConverseNow Founder and CEO Vinay Shukla told me that the company’s assistants are currently live at 750 restaurant locations, that they achieve an 85 percent order accuracy rate, and that they increase check size by 25 percent.

But ConverseNow is a much bigger play than simply automating the drive-thru. The company is developing a suite of tools to automate order-taking in many different restaurant scenarios. Shukla said he sees ConverseNow as like a Twilio for restaurants, providing one artificially intelligent glue that can power ordering via drive-thru, mobile app, phone, kiosk, etc.

Shukla told me that while ConverseNow can save on labor costs for a restaurant, it’s also being developed with the restaurant operator in mind. “Nobody is talking about the operator experience,” Shukla said, “[Customers] get frustrated, which is causing a lot of attrition. The experience for the operator at the store is not great.”

We have steadily been seeing more adoption of AI at QSRs over the past couple of years, especially in the drive-thru. A company called 5Thru is connecting license plates with customer profiles to provide order history as well as upsell recommendations (KFC was at one point considering adopting this type of service). Valyant AI is another startup that has implemented voice-based ordering for QSRs. More recently, last month McDonald’s said it was testing AI-based drive-thrus based on Apprente (which McDonald’s acquired in 2019) at 10 Chicago locations.

Shukla said it will use the new funding to further develop and scale up its products.

June 3, 2021

McDonald’s Testing AI-Powered Drive-Thrus in Chicago

McDonald’s has started testing out drive-thrus that use artificial intelligence systems, rather than humans, to take orders. CNBC reported yesterday that the new automated drive-thrus are in use at 10 Chicago McDonald’s locations.

The new system is based on the voice platform built by Apprente, which McDonald’s acquired in 2019. According to McDonald’s, restaurants using the system are seeing an 85 percent order accuracy rate, with only about one-fifth of orders requiring human intervention.

AI-powered drive-thrus can reduce customer wait times and allow restaurants to shift its in-store workforce. A computer that understands natural language is always on and available to take orders. It could also be tied in with other automated systems that know a customer’s purchasing history to automatically make recommendations. With improved understanding accuracy, a restaurant would no longer need a dedicated person to take (or confirm) a drive-thru order, allowing more people to do more customer service or expedite orders.

McDonald’s CEO Chris Kempczinski told Alliance Bernstein’s Strategic Decisions conference that a big issue ahead for the AI-powered drive-thru is scaling. CNBC reports Kempczinski as saying “Now there’s a big leap from going to 10 restaurants in Chicago to 14,000 restaurants across the U.S., with an infinite number of promo permutations, menu permutations, dialect permutations, weather — and on and on and on.”

The Apprente acquisition appears to be working out better than Dynamic Yield, which McDonald’s also acquired in 2019. Dynamic Yield generated automated menu recommendations based on factors like weather, and was supposed to be integrated into self-service kiosks and drive-thrus as well. However, this tech didn’t yield the results McDonald’s was looking for and in March of this year The Wall Street Journal reported McDonald’s was looking to sell part of Dynamic Yield.

While McDonald’s Apprente acquisition may have pre-dated the pandemic, last year certainly accelerated the need for enhanced drive-thru technology as dining rooms were forced to shut down. In a February 2021 survey, BlueDot reported that 91 percent of respondents said they had visited drive-thrus the previous month and that long wait times were a “dealbreaker.” Most major QSRs have been doubling down on their drive-thru capabilities to meet this demand, adding capacity and building restaurants around takeout rather than dine-in.

In addition to adding AI assistants, McDonald’s has previously said that it will add other features to its drive-thru such as express lanes for digital orders and conveyor belts to carry food out to customers.

Kempczinski also told the conference that McDonald’s is also exploring ways to automate parts of the kitchen such as the grill or fryer. However he said any such move in the back of the house is still a more than five years out as the technology is too expensive right now.

May 30, 2021

A Few Thoughts on Yum’s Dragontail Deal

This is the web version of our newsletter. Sign up today to get updates on the rapidly changing nature of the food tech industry.

Greetings from Eastern Kentucky, where I’m taking a quick breather from the restaurant world and am off to tour AppHarvest’s high-tech greenhouse.  

But first, a few quick thoughts on the recent Yum! Brands acquisition of Dragontail Systems, which was announced at the end of last week. In case it wasn’t clear already, Yum — parent company of Taco Bell, KFC, Pizza Hut, and The Habit Burger Grill — has big ambitions when it comes to its technology plans. The Dragontail deal suggests that a big chunk of that ambition is to enable better delivery at its restaurants. 

Australia-based Dragontail Systems is known around the restaurant biz for its AI-based workflow management system for kitchens. The system can process a restaurant’s incoming orders, manage each food item’s cook time so that it’s ready when the delivery driver arrives (not before or after), assist with managing and scheduling drivers, and provide real-time updates. In the past, restaurants have also used Dragontail’s AI-based camera to assess food quality and ensure safety and cleanliness standards in the kitchen. 

Dragontail is Yum’s third major technology acquisition so far in 2021. Earlier in the year, Yum acquired AI firm Kvantum to assist with its marketing campaign analytics, and Tictuk, which makes omnichannel ordering software that lets consumers place orders via social media, SMS, email, and other formats.

The Dragontail Acquisition, is hyper-focused on improving delivery operations, and pizza delivery in particular. Yum said it already has Dragontail installed in roughly 1,500 Pizza Hut stores across 10 countries. If a wider implementation of the tech is successful, Yum will eventually install the tech further around its Pizza Hut locations, and eventually across all its brands. 

Of course, Dragontail originally debuted its system at a Domino’s location in Australia, so Yum’s bid to buy the company is at least in part a way of competing. I doubt fighting Domino’s is the only motivator, though, as the deal seems aimed at a larger move to make food preparation more efficient at Yum restaurants.

Restaurant kitchens have always operated in a kind of orchestrated chaos, but with many businesses still struggling in the wake of last year’s events, efficiency is everyone’s favorite word right now. On paper, Dragontail’s capabilities sound easy: managing drivers’ arrivals and departures, determining more precise fire times for each food item, assessing the quality of the food more precisely. But as anyone who’s ever worked in a restaurant kitchen knows, getting one detail wrong can set off an entire chain of mishaps that lead to lower-quality food, unhappy customers and, in the end, lost money. With the Dragontail acquisition, Yum is placing its bets on tech to help the company avoid that scenario. 

Plant Power Fast Food Chain Raises $7.5M Series A – The San Diego, California-based all-vegan chain trying to redefine the concept of “fast” food said it will use its new funding to continue its expansion plans and focus on new corporate unit development.

7Shifts Raises $21.5M for Its Workforce Management Platform for Restaurants – The company will use its $21.5 million Series B round to add new features that simplify the process of managing, scheduling, and communicating with restaurant workers.

Square’s Back-of-House Display System Now Available for Delivery-Only Restaurants – Payments company Square recently made its restaurant software stack a little more versatile when it launched its Square KDS back-of-house display system as a standalone item available via subscription.

May 4, 2021

Miso Robotics Launches CookRight to Automate Restaurant Cooking (Without the Robot)

Miso Robotics today announced CookRight, a new software product for restaurants that delivers the artificial intelligence (AI)-powered cooking of its Flippy robot — without the robotic hardware.

The new CookRight platform uses cameras mounted above a grill along with a tablet computer. The cameras look down at the food being prepared and use a combination of computer vision, thermal detection and AI to identify a cut of meat, its thickness and cooking progress. It’s the same software system used by Flippy, only instead of a providing instructions to a robotic arm, its guides a human cook. A display on the accompanying tablet shows what CookRight is “seeing” and gives instructions for each item on the grill. Miso says that CookRight currently works with burgers, chicken, fish, steak, sausage, hot dogs, and more.

Miso Robotics Chairman and President, Buck Jordan, told me by phone last week that using CookRight can help reduce foodborne illnesses by ensuring food is cooked properly, and with the guided cooking, turn just about any cook into an expert griller with no extensive culinary training.

Additionally, CookRight integrates with a restaurant’s POS and ordering systems to automate coordination of meal prep. So if an order comes in for delivery through an app like GrubHub, CookRight will know when the driver is on their way and be able to time the grilling accordingly.

With CookRight, Miso is aiming to expand its market base with a lower cost automated cooking solution than Flippy. Though the price of Flippy has steadily come down from $60,000 upfront to $2,000 a month, that’s still a lot of money for hardware that needs to be installed in existing kitchen spaces, maintained and have workers be trained on. Additionally, smaller mom n’ pop restaurants may not have the space necessary for a Flippy installation. The biggest customer for Flippy at this point has been White Castle, which will be operating the robot at 11 of its locations.

With its minimal hardware setup and a subscription cost of $100 a month, CookRight is a much more affordable option for smaller and medium-sized restaurants looking to automate some of their processes. This, in turn, will allow Miso to scale to more restaurants more quickly.

There are other players in the restaurant tech space like Dragontail Systems, which uses AI to coordinate and optimize order workflow in the kitchen. But with CookRight, Miso is taking that automated integration and optimization a step further into the actual cooking of the food.

If you want to see the future of automated cooking, then you should definitely check out Articulate, our food robotics and automation virtual summit happening on May 18. Buck Jordan will be speaking as will execs from restaurant robot companies like Karakuri, Blendid, Piestro and Mukunda Foods. Get your ticket today!

April 3, 2021

Food Tech News: Google AI Cake, Chipotle’s Bitcoin Giveaway and Robot Food Delivery

Welcome to the first Food Tech News round-up of April! This week we have news on a cake created by Google artificial intelligence, Kiwibot hitting the streets of Santa Monica, Ember’s new travel charger, and Chipotle’s bitcoin giveaway.

Google artificial intelligence created a cake recipe in partnership with Mars Wrigley

Google Cloud engineers created a machine learning model that uses hundreds of existing baked good recipes to develop a completely new recipe. The result was a “Cakie” (a cake and cookie hybrid) and the components of what makes a cake and cookie were generated with artificial intelligence. For the partnership with Mars Wrigley UK, Maltesers (chocolate-covered malt balls) were incorporated into the recipe to create the first-ever “Maltesers AI Cake.” Google trends revealed that “sweet and salty” was a top search trend, and the cake recipe used a buttercream frosting infused with Marmite. Earlier this year, the same machine learning model was used to create two totally new baked good recipes, the “Cakie” and “Breakie”.

Kiwibot and MealMe partner for food delivery in Santa Monica

MealMe, an app that compares prices and times of food delivery services, and Kiwibot, a teleoperated robot service that delivers food, have partnered to deliver food in Santa Monica. On April 1st, the companies began delivery for Blue Plate Taco and Red O Restaurant on Ocean Ave. Kiwibot’s robots provide contactless food delivery, and so far 100k deliveries have been completed in Berkeley, Los Angeles, San Jose, Denver, Taipei, and Medellin,

Ember launches car charger to keep beverages warm on the road

Ember, the creator of the self-heating coffee mug, has created a car charger to keep Ember Travel Mugs warm all day. The Ember Travel Mug is capable of keeping a beverage warm for three hours, but now it can be plugged directly into a car charger for an all-day charge. The car charger costs $49.95 on Ember’s website.

Photo from Chipotle

Chipotle hosted a giveaway of free burritos and bitcoin

For National Burrito Day, Chipotle partnered with the founder of Coil, Stefan Thomas, to giveaway $100,000 worth of Bitcoin and $100,000 in burritos. To win, contestants had 10 chances to guess a six-digit code. The giveaway only lasted for nine hours on April 1st on BurritosOrBitcoin.com. Update: I tried, and did not win.

March 29, 2021

Food Tech Show Live: Sony Invests in our Robot Chef Future

The Spoon team recently got together on Clubhouse to talk about some of the most interesting food tech and future food stories of the week. This time around, we were also joined by food tech investor Brian Frank.

If you’d like to join us for the live recording, make sure to follow The Spoon’s Food Tech Live club on Clubhouse, where you’ll find us recording our weekly news review every Friday.

The stories we talked about this week include:

  • Cell-Cultured Fish Startup Bluu Biosciences Raises €7 million
  • The Rise of ‘Premium’ Cultured Meat Startups
  • Sony Invests in Analytical Flavor Systems and our Robot Chef Future
  • NASA Harvest Partners with CropX to Combine Soil Monitoring and Satellite Data
  • Ex-WeWorkers Launching Santa, A Hybrid ‘Retail Experience’ Startup Focused on ‘Small US Cities’

As always, you can find the Food Tech Show on Apple Podcasts, Spotify or wherever you listen to podcasts. You can also download direct or just click play below.

March 24, 2021

Sony Invests in Analytical Flavor Systems To Help Our Robot Chefs of the Future Better Predict What We Want to Eat

When Analytical Flavor Systems (AFS) CEO Jason Cohen talks about his company’s technology, it usually involves a story about helping a big food company. In particular, how AFS’s AI flavor platform, Gastrograph AI, can predict how a new product such as a bag of chips or energy drink might perform in a new country or with a new demographic group.

Ask him to tell you about his company’s flagship product in the future, though, and the conversation might well include discussing how it helped a robot chef decide what exactly to make us for dinner.

That’s because the New York-based startup has recently taken a corporate investment round from Sony, which late last year announced a dedicated project focused specifically on developing food focused AI and robotics technology.

According to Cohen, Sony’s investment team believes AFS’s technology could help them achieve their goal of creating robot chefs that know exactly what to cook.

“Once you get a food robot up and running, you don’t know what people are gonna want to eat, what they’re going to prefer, like and dislike,” said Cohen in a Zoom call.

But with a technology like Gastrograph, robots of the future might be able to more accurately predict flavor combinations consumers better than humans themselves.

“Consumers have no idea what they like and dislike,” said Cohen. “They’re actually really bad at determining whether, say, they want more vanilla in their vanilla cup. You ask them and they say ‘yes’, you put it down and they want less. Sony recognized that and that’s what part of this investment is for: future innovation and growth, to help them accomplish their goals of getting cooking robots into every home.”

In addition to Sony, AFS also took on corporate investment from BASF. According to Cohen, the German conglomerate’s plant breeding and genomics program is interested in using Gastrograph AI to help them in seed development so they can ultimately predict with a higher success rate with the crops they are breeding.

“It takes a really long time, anywhere from a year to a couple of years, even decades, on some crops, in order to breed in specific traits,” said Cohen. “And so they work with us to rapidly screen and profile the fruits that they are market leaders in, or the fruits that they want to become market leaders in, and to breed better seed stock that they can then sell to farmers in different countries around the world.”

While Cohen wouldn’t disclose the amount of investment from Sony and BASF, he did say that this was an in-between “corporate”round that they had been discussing with the two strategic investors before COVID hit, so when everything shut down in March of last year things were “already in place.” Next year, said Cohen, he expects the company will raise a Series A round of funding.

Speaking of COVID, I asked Cohen how his company did this past year, and he said their core business of working with CPG innovation teams actually benefitted because as companies wanted to keep the innovation pipelines moving during a time of severe travel restrictions and AFS’s technology allowed them to do that. Since the normal big food practice of testing products in-market with focus groups curtailed during lockdowns, some companies relied on Gastrograph AI’s technology to predict how a new product might do.

This recent investment follows a 2018 seed round investment of $4 million.

March 3, 2021

HungerRush Launches Its AI-based Text-to-Order Tech for Restaurant Chains

Restaurant management software company HungerRush announced today that it launched its artificial intelligence-driven text-to-order product.

Dubbed HungerRush TextAI, the new feature uses natural language processing to interpret and place orders that come in via SMS. The new product is the result of HungerRush’s acquisition of OrdrAI last December. In its press announcement today, HungerRush said that TextAI works better than traditional phone or other employee-assisted formats by virtually eliminating order errors.

Text AI also integrates with HungerRush’s overall suite of products that include POS, delivery management, online ordering, mobile app, reporting and management, payment processing, and loyalty program.

HungerRush’s TextAI product launch comes at an interesting moment for AI integration for restaurants. Artificial Intelligence-driven software grabbed headlines last year as Kea, Clinc and even Google all had products that used natural language processing to hold “conversations” with restaurant customers. The promise, of course is that AI can take over incoming customer orders, allowing humans to perform more food creation and order expediting.

But in a signal that AI might not work as well as had been predicted, news broke this week that McDonald’s may sell Dynamic Yield, the AI startup it acquired in 2019. Granted, Dynamic Yield’s implementation of AI is more about creating dynamic menus that improved upselling (e.g., it’s cold outside, the menu highlighted warm beverages), so it’s a bit apples to oranges when compared with text-to-order. But if McDonald’s reportedly couldn’t even get a 1 percent bump from its implementation of AI, that could impact how other restaurant chains perceive the overall value of AI.

That’s not to say restaurants will abandon AI altogether. All of these technologies are new, and, not for nothing, are being tested under the unusual conditions of a pandemic, which has us ordering more takeout and delivery. Restaurants, which are seeing diminished value in third-party delivery services, need to save money where they can. If AI-driven text-to-order can drive sales and help restaurants get more value out of their labor, then that will go a long way towards adoption.

Previous
Next

Primary Sidebar

Footer

  • About
  • Sponsor the Spoon
  • The Spoon Events
  • Spoon Plus

© 2016–2025 The Spoon. All rights reserved.

  • Facebook
  • Instagram
  • LinkedIn
  • RSS
  • Twitter
  • YouTube
 

Loading Comments...