Close Ad

The Spoon

Daily news and analysis about the food tech revolution

robots

Researchers Create Simulator to Help Robots Wield Knives

Robotics researchers from NVIDIA and the University of Southern California (USC) announced today the first differentiable simulator for robotic cutting, or DiSECt for short. This new simulator can predict forces that will act on the knife as it pushes and slices through soft materials like fruits and vegetables.

Your first reaction might be, why do they need all that simulator science when you can just install a sharp blade on a robotic arm and smash it down? That’s certainly one solution, but part of the reason robot researchers like NVIDIA, and Sony and Panasonic all work with food is because food is oddly-shaped, has different textures and is delicate. If a robot can successfully work with soft objects like food, it can carry those techniques over to other applications like surgery (where plunging knives down is frowned upon).

Cutting through food with precision and care is actually quite complex. It requires feedback, adaptation, motion control and parameter setting as the knife makes its way through the object. Additionally, since each piece of fruit or vegetable is unique, the robot needs to adjust its cutting with each new object.

NVIDIA shared with us an advanced look at an article explaining the DiSECt research that was recently presented at 2021 Robotics: Science and Systems (RSS) conference. I’m not going to lie, it is dense and jargon heavy with paragraphs like this:

DiSECt implements the commonly used Finite Element Method (FEM) to simulate deformable materials, such as foodstuffs. The object to be cut is represented by a 3D mesh which consists of tetrahedral elements. Along the cutting surface we slice the mesh following the Virtual Node Algorithm [4]. This algorithm duplicates the mesh elements that intersect the cutting surface, and adds additional, so-called “virtual” vertices on the edges where these elements are cut. The virtual nodes add extra degrees of freedom to accurately simulate the contact dynamics of the knife when it presses and slices through the mesh.

But rather than focusing on the specifics of the research, there are some broader takeaways anyone in food tech can appreciate. First, DiSECt illustrates the continued importance of simulation and synthetic data in training robots. NVIDIA has actually built a kitchen as a training ground for its robots where it uses synthetic data and computerized simulation to virtually teach a robot tasks like identifying and picking up a box of Cheeze-Its. Similarly, DiSECt trains a robot to use a knife through simulation first, which can then be applied to the cutting object in the real world.

Additionally, giving robots added abilities will make them more useful in taking over dangerous tasks like repetitive cutting. Right now, robots in restaurants are frying foods and even making pizzas, but they aren’t doing more highly skilled, precision tasks such as cutting and slicing. A robot can’t get injured while cutting and could bring more safety to restaurant kitchens.

The good news for those interested in this type of cutting-edge research is that NVIDIA and USC are not the only companies doing work in this field. In 2019, researchers from Iowa State University published a similar paper on the intricacies of robot slicing.

DiSECt: A Differentiable Simulation Enginefor Autonomous Robotic Cutting

Rise of the Beer Robots!

This is the web version of our Weekly Spoon newsletter. Subscribe today to get all the best food tech news delivered direct to your inbox!

Beer is the pizza of beverage automation.

Previously, I’ve noted that if you want to see the future of food tech, you should look at pizza. From robot assembly to self-driving vehicle (and drone!) delivery to vending machines, pizza tends to be a vanguard for innovation.

Just as pizza paves the way for interesting food technology, beer appears to be the beverage to watch when you want to know where drink automation is heading.

Consider two news bits we highlighted on The Spoon this week:

Part of what makes these two stories of note is that the companies featured are on opposite ends of the spectrum. Hop Robotics is a small, one-man operation that only has one commercial prototype; Heineken is a giant global using robotics basically as a promotional gimmick. (You can enter to win its Beer Outdoor Transporter.) Regardless of their intentions, both companies are looking to make it easier and faster for you to grab a brew.

Hop and Heineken aren’t alone in this endeavor to speed up beer service, as we’ve steadily seen other automation startups come to market. In the U.K., EBar makes big, mobile vending machines meant for events and large venues that will pour a pint in under 30 seconds. The Revolmatic comes out of Poland, and is a countertop machine with a rotating tray that can dispense 450 beers an hour. Macco Robotics in Spain is taking a slightly different approach, employing a humanoid robot with arms to pour your beer in 23 seconds.

All of these machines are meant for events and large gatherings (sports, concerts, conferences) and, when set up in age-restricted areas, can act as unattended beer retail operations. This is helpful in a couple of ways. First, these robots can take over the grunt work of just pouring hundreds of beers an hour. These speedy workhorse machines can help customers spend less time in line for drinks and more time at their event. Automation can reduce overages with consistent, perfectly portioned pours, saving money and reducing waste. These robots also free up human bartenders to focus on more complicated drinks and customer service. And finally, while we are coming out of this pandemic (fingers crossed), venues and retailers will still be looking for and adopting contactless technologies that reduce human-to-human interaction.

Beer isn’t the only beverage getting the automation treatment. Both Rotender and Celia robots are making mixed cocktails, and Botrista just raised $10 million this week for its cloud-connected mocktail and fusion drink dispenser. But I think we’ll see the most automation activity in the beer space. I mean, beer is a huge market. According to the National Beer Wholesalers Association, 2020 U.S. retail sales of beer and malt-based beverages was $100 billion, and that was during a pandemic, when restaurants, bars and stadiums were closed. (Sales were $120 billion in 2019.) And for automation startups, a beer machine is just easier to make because it doesn’t require as many mechanical bits and bobs that are needed to make cocktails (ice, different bottles of booze, mixers, etc.).

Perhaps what’s most fun about Hop Robotics is its size. If one guy in South Carolina can build a working beer robot in his garage, imagine what well-funded companies will create. Whatever the future brings, beer and pizza night will never be the same.

Image via Vegano.

More Headlines

Vegano Launches an All-Vegan E-Commerce Grocery Marketplace in Canada – For now, the service operates in the Metro Vancouver area as well as Squamish and Whistler. The company said it plans to expand to Toronto and Montreal by the end of this year in addition to heading Stateside and launching in Los Angeles.

Czech Online Grocer Rohlik Raises $119M, Its Second Nine-Digit Round This Year – This funding comes just months after Rohlik raised a €190 million (~$230 million USD) Series B round in March of this year. This brings the total amount of funding raised by Rohlik to nearly $380 million.

AiFi and Trigo CEOs Weigh in on When Cashierless Checkout Will Go Mainstream – TL;DR, look for most major cities to have at least one next year, with more saturation coming in ten years.

Farm.One Launches a New Vertical Farming Facility in Brooklyn – The space will grow various microgreens as well as herbs and some flowers. All crops are grown using the hydroponic method and artificial lighting, with plants harvested “hours before delivery,” according to the company.

Heineken B.O.T. Follows You Around With Ice Cold Beer in Tow

I live in the Pacific Northwest which, as you may have heard, is experiencing a record-breaking heatwave. It is hot, especially for a region that lives under cloud cover for most of the year. What we really could have used up here over the past few days is the Heineken B.O.T. (Beer Outdoor Transporter), an autonomous little robot that will literally follow you around carrying a cooler full of ice cold beer (or whatever).

Hat tip to Hype Beast for bringing the B.O.T. to our attention. The robot is autonomous and uses motion sensors to follow you around and avoid obstacles. As the promotional video points out, lugging coolers around on a hot day is heavy and no fun. The B.O.T.’s cooler holds 12 cans of beer and ice and since it goes where you go, a frosty beverage is always within arm’s reach.

Heineken B.O.T. (Beer Outdoor Transporter)

The idea of autonomous robots bringing refreshments to people isn’t new. A couple years back, Pepsi was using Robby robots to being snacks to students at the University of the Pacific in Stockton. More recently, Cheetah Mobile in China launched FANBOT, which is like a mobile bodega — though it’s used indoors in places like hotels and malls. And Yo-Kai Express is currently developing an autonomous ramen vending machine for students to hail on college campuses.

But don’t expect to see Heineken B.O.T.s scurrying around public sidewalks anytime soon, however, as this appears to be a promotional play from Heineken. It’s holding a contest in July where you can enter to win your own B.O.T. and stay refreshed through the hot summer months.

Soft Robotics Raises $10M to Add 3D Vision and AI to its Octopus-like Grippers

Soft Robotics, which is best known for making octopus-like grippers for robots, announced today that it has raised a $10 million extension to the $23 million Series B round it raised in January 2020. The round was co-led by Material Impact, Scale Venture Partners and Calibrate Ventures, and adds Tyson Ventures (the venture arm of Tyson Foods) to the syndicate. ABB Technology Ventures and Tekfen Ventures participated as well. This brings the total amount of funding raised by Soft Robotics to $58 million.

Soft Robotics uses rubber tipped grippers with “air actuated soft elastomeric end effectors” that mimic an octopus, allowing robotic arms to pick up odd-shaped and delicate items like eggs and bread without crushing them. The company says the new capital will help Soft Robotics launch its new SoftAI technology, which adds layers of 3D vision and artificial intelligence to its gripping solution.

According to Soft Robotics’ website, “SoftAI will evaluate the pick scene and automatically choose the best grasp and ideal robot trajectory to optimize rate and reduce product damage.” It’s easy to see how this type of automated discernment would come in handy for a company like Tyson Foods (which was already using Soft Robotics before it invested), which needs to pick up and pack all different types of animal products of varying shapes and sizes.

Chicken Wing and Poultry Automation with mGripAI

In addition to its new technology, Soft Robotics said its new funding will go towards commercial expansion to keep up with pandemic-driven demand. Last year COVID-19 exposed shortcomings in our food supply chain, with meatpacking facilities, which were already a dangerous place to work, becoming hot spots for the virus. Implementing robots in a meatpacking or other food-related factory can help add additional safety and social distancing to the work environment. Robotic arms can work all day without fatigue or injury, and placing robots on a line can help space out workers, so people aren’t working right next to each other.

During our first ArituclATE food conference back in 2019, a robotics researcher told me that robotic “grippers all suck.” But that appears to be changing. In addition to Soft Robotics’ octopus approach, new technologies based on origami (paper folding) and kirigami (paper cutting) are creating entirely new types of gripping technology that can be used for odd-shaped and delicate items. The combination of the pandemic and investor interest could help fuel accelerated development and implementation of this new gripper technology and unlock new areas and uses for robots in food production.

2021 Restaurant Tech EcoSystem: Serving Up a Digital Lifeline

In collaboration with TechTable and Culterra Capital, we are pleased to share an updated 2021 Restaurant Tech Ecosystem map, sponsored by Back of House, a community of restaurateurs to find and share top-reviewed tech solutions. Download the map here. 

It is an understatement to say that the restaurant industry went through a massive shift since we published our 2019 Restaurant Tech Ecosystem map. The pandemic’s economic toll on the industry has been grave, though notably, the toll was not evenly distributed. A higher level of digital maturity was a clear success indicator for most restaurants that survived the crisis (as detailed by McKinsey here). 

Thus, in a year that was challenging for all, we did find one bright spot from the pandemic: many of our past predictions around tech adoption were significantly accelerated, shrinking from years to months. 

In fact, in the past 18 months, technology solutions across the restaurant and hospitality industry evolved at such a fast pace that keeping up with changes proved challenging, even for those of us who work in the space. This rapid rate of adoption in the industry caused even the technophobes in hospitality to rapidly embrace tech solutions. 

The most notable growth areas were in the areas of Ordering/Delivery and On-Premise Ordering/Payments Tech, including kiosks, mobile ordering and payments, and cashierless checkout. In addition to the acrobatic feats from restaurant operators, we also saw tech companies reinventing themselves to stay in business during the pandemic.

With that in mind, we are pleased to share our 2021 Restaurant Tech Ecosystem, which serves as a current heat map of the broader ecosystem (and is clearly not exhaustive). 

Click to Enlarge

In order to help operators, entrepreneurs and investors continue to understand and digest this quickly evolving landscape, we also highlight some of the essential shifts and sector themes below, plus a few predictions for the year to come.

Help Wanted

Finding and securing hospitality staff has never been so challenging. As thousands of hospitality workers were left unemployed by the pandemic, or stymied by the risks of the frontlines, many have moved on to find work in other fields, leaving a huge gap in talent. 

As a result, the urgency to leverage robotics, automation, computer vision, and voice technologies will continue to increase as the hospitality industry aims to do more with less staff. And while discussions around robots within hospitality have always been cautious — because we don’t want to put people out of work — we believe we will continue to see more opportunities in the near-term for human-assisting robots (versus human replacement by robots). 

For example, as restaurant operators seek to offset workforce challenges, there are numerous opportunities for specific task automation of repetitive, dangerous or mundane tasks like dishwashing, precision preparation/cooking, food waste management, bar/food inventory, and quality control.

Another area ripe for automation via AI-driven voice tech includes drive-thrus and digital ordering. When people think of a traditional drive-thru, they likely picture a garbled voice and screaming the order into a speaker, hoping their order is correct. But we are seeing many of the larger chains replacing human voices with automated voice assistants to speed up service, order accuracy, and upsell rates. We’ve seen estimates that drive-thru automation can reduce customer wait time by 10 to 25 percent, which is compelling given that the former CEO of McDonalds previously declared that for every six seconds saved at a drive-thru is equal to an increase of 1 percent in sales.

Ghost Kitchens: It’s Complicated

While many restaurant operators were broadly familiar with the concept of ghost kitchens and virtual brands before the pandemic, these formats are now prevalent in most discussions on the burgeoning post-pandemic restaurant industry. 

Whether part of an existing kitchen or a separate commissary kitchen, the ghost kitchen’s purpose is to fulfill online orders for delivery or pick up. Ghost kitchens have the potential to solve real challenges for their restaurant customers, and there are tremendous variations on the economics, setup, and ideal use cases.

So then what’s so complicated about ghost kitchens? 

The rapid growth in consumer demand for restaurant delivery and the high usage of third-party ordering/delivery apps pushed restaurant operators to explore different avenues to expand their access points and footprint beyond their existing restaurants. 

However success (a.k.a. profitability) within the confines of a ghost kitchen business model is primarily driven by volume of daily orders, average order value, and percentage of direct channel sales versus third-party sales. This is why ghost kitchens are primarily well-suited for larger brands, as most local restaurants simply do not meet the average requirements to warrant a ghost kitchen endeavor. (If you are curious to crunch the numbers, check out this excellent ghost kitchen calculator created by Kitchen Fund.)

Further, the lines are beginning to blur between delivery and ghost kitchen platforms. We are entering a world where these platforms are increasingly supporting their own virtual brands and/or next-gen food courts, oftentimes by using the ordering/menu data captured from current restaurants using their platform. Thus local operators will be battling for market share against larger chains which are using ghost kitchens to extend their reach and volume, as well as additional competition from ghost kitchen platforms themselves.

Enter the The Mobile-Only Experience

Stateside, we’ve increasingly been adopting mobile-first ordering and marketing strategies, but the mobile-only approach (often seen in Asia) wasn’t widely embraced before the pandemic. Now, whether via QR codes, apps or mobile web, there has been a huge shift towards mobile-optimized menus, ordering and payments which eliminate or reduce most employee/customer contact. This can help to improve the guest experience via increased speed and fewer errors. For fine dining, this also saves time/costs in printing and sourcing supplies for paper menus. 

Successful operators will prioritize their tech strategy to capture as much digital data as possible in order to personalize offers, segment customers and influence behavior, and a mobile-first/mobile-only approach creates a compelling opportunity to increase both first-party and third-party data capture.

As it can be dizzying for operators to decide how to best leverage their digital data, we predict high growth for the tech partners which are helping operators utilize customer data to better uphold their brand, funnel customers into more profitable channels, and make better decisions about merchandising, pricing, and promotions.

Aggregators Will Continue to Disrupt the Customer Journey

While many of the technologies we discuss here are more operational, we also want to address the customer search and discovery experience. While this was an important topic in pre-COVID times, it is all the more so now, when disruptions and uncertainties are pushing customers to regularly search for what nearby food options are actually open, and whether they offer delivery, curbside, or takeout options. 

We have also reached a point where Google/Google Maps have become the defacto top of the funnel for a majority of restaurant consumers. Thus it is increasingly critical for restaurant operators to proactively manage their full digital footprint, and provide up-to-date information that customers can trust — especially across all third party platforms. 

For example, even though Google profiles include a link to a restaurant’s own website, that little link is eclipsed by the amount of ad-driven real estate that the third-party aggregators/marketplaces get within each profile. 

Growth Categories to Watch in 2022

  • Voice / Bot Technology
  • Robotics / Automation
  • Shared / Ghost Kitchens
  • Food Safety / Quality (a new category for the 2021 map)
  • Ordering and Payments will continue to evolve
  • Marketing Analytics / CRM, and Order / Delivery (both B2B and consumer-facing marketplaces) will continue to consolidate.

As always, we welcome your thoughts and reactions, and look forward to continuing to follow this sector together in the coming years.

Hear Brita and other restaurant tech leaders at The Spoon’s Restaurant Tech Virtual Summit on August 17th. A limited number of complimentary tickets are available, so register today!

Grab Launched AHBOI Robot to Cover Middle-Meter Meal Delivery in a Singapore Mall

We’ve covered “middle mile” deliveries before, but a new robot in Singapore is covering what we’ll dub the “middle meter.” Last week, Singapore-based delivery service Grab launched its AHBOI robot at the Payar Lebar Quarters (PLQ) mall to speed up delivery orders placed at the restaurants located inside the shopping center.

AHBOI, which stands for Autonomous Handling and Batching Operating Intelligence, is a large, green rectangular mobile locker that acts as a food expediter. According to a Grab corporate blog post, the AHBOI collects delivery orders from different restaurants inside the mall and autonomously brings them to a central collection point where delivery drivers pick them up for the last mile delivery.

By shuttling food from restaurants across these middle meters to central pickup point, Grab says it can shave as much as 15 minutes off of a delivery time because the delivery driver doesn’t need to navigate their way to and inside the mall.

We are starting to see more automation appear over the middle meters and miles for food delivery. QSRs like Burger King and McDonald’s are eyeing conveyor belts at restaurants to speed up the movement of food from the kitchen to drive-thru customers. And companies like Valqari are setting up smart lockers where drones can drop off meals at a central location for delivery drivers or customers to pick up from. Using robots to consolidate orders in a large indoor setting like a mall makes a lot of sense.

As part of its blog post, Grab encouraged users to place more orders from the mall that would use AHBOI. The idea being that the more the robot traverses PLQ, the more AHBOI will learn about its environment and the better the robot will get at driving itself around the mall. Of course, more orders also means more revenue for Grab, which they could use to make more robots.

Botrista Raises $10M Series A for Its DrinkBot Automated Drink Dispenser

Botrista, the company behind the commercial DrinkBot automated beverage dispenser, announced today that it has raised $10 million in Series A funding. The round was led by Purestone Capital and La Kaffa International with participation from Sony Innovation Fund, Middleby Corporation and PIDC. This brings the total amount of funding raised by Botrista to $16 million.

Meant for restaurants an other foodservice companies, the DrinkBot is a cloud-connected automated drink maker, dispensing mocktails, infused teas and lattes, iced coffees, lemonades and more without the need for a full bar. Drinks are ordered via an on-board touchscreen, so the experience is contactless, and they are mixed and served in less than 20 seconds.

Botrista is taking a vertically integrated approach as it comes to market. The company provides the hardware for free, charging a monthly maintenance fee and selling the drink ingredients, which as of last year were $1.40 – $1.90 per drink. DrinkBots connect to the Botrista CloudBar for drink recipes, automated inventory management, as well as sales and menu performance analytics.

In a press release sent to The Spoon, Botrista said it experienced 10x growth year-over-year, which isn’t that hard to believe. The pandemic is driving demand for more contactless food and beverage preparation, as well as the need for takeout and delivery-related tech. The Botrista can churn out drinks continuously, quickly and at the touch of a button — perfect for high volume establishments like ghost kitchens and restaurants with high off-premises volume.

Automated mocktail and juice dispensing is becoming a hot little sub-sector of the food and beverage robotic space. Over in Switzerland, Smyze’s robot baristas also make a bevy of juice beverages. And last year, SomaBar, which makes a countertop drink dispenser, pivoted away from Soju-based cocktails to have its machine create juices, teas and regular mixed drinks.

It’s worth noting that both Sony and Middleby participated in Botrista’s latest funding. Last year Sony set up an artificial intelligence unit to work on both recipe creation and robotics. It’s easy to see how DrinkBot’s combination of data and automation fits in with that endeavor. And Middleby, a giant in foodservice automation, could open up a large network of customers for Botrista.

Botrista said it will use its new funding to scale up deployment operations to roll out DrinkBot nationally.

Print a Drink 3D Prints Designs Inside a Cocktail, Develops Smaller Machine for Corporations

We’ve seen 3D printers create cake decorations, personalized vitamins, and even cultured beef. And now, thanks to Print a Drink’s robot, we’ve seen custom designs printed inside a cocktail. You might think such beverage witchcraft would be impossible. I mean, how could a design be suspended and hold its shape in anything other than a jello shot? Turns out it just takes the right drink, the right droplet and the precision of a robotic arm.

Based in Austria, Print a Drink has actually been around for three years. It was started by Benjamin Greimel as a university research project. Since that time, Print a Drink has created two working robots (one in the U.S. and one in Europe) that up until the pandemic would travel to special events and conferences printing out custom designs inside drinks at parties and such.

So how does it work? Print a Drink uses a robotic arm with a custom-made printer head attached to it. The robot uses a glass needle to inject a food-grade, oil-based liquid inside a drink. The drink itself needs to be less than 40 percent alcohol and can’t be a straight shot of something like vodka or whiskey because the injected beads won’t hold and will float to the surface. Greimel explained to me via video chat this week that the combination of liquid density, temperature and robotic movement allow the designs to last for roughly 10 minutes before dissipating.

Coordinating all those puzzle pieces is complicated to say the least. In addition to setting up the robot at an event and operating it, there are specific requirements around drinks that can be used, and designs need to be uploaded into the robot. Plus, there are safety concerns because the robotic arm does move about pretty quickly. Because of all those reasons, Print a Drink’s business has been around renting the robot ($2,500 – $5,000, depending on the event) and not selling them outright. In addition to all of the complications above, staff would need to be trained properly on how to use the machine, and chances are good that the people operating the devices are not roboticists who can troubleshoot.

To make Print a Drink more accessible, Greimel and his partner (the only two people at the company) have developed a smaller, self-contained version of the robot that is roughly the size of a countertop coffee machine. But don’t expect a consumer version for your next backyard soirée. This smaller version is still complicated, and still requires training, so the company is targeting large corporations like Disney or a hotel chain like Hilton where it could be installed and used for special events or promotions. Greimel said the first prototype of this smaller Print a Drink will be available in the next week.

Though more specialized, Print a Drink is part of a bigger automation movement happening with booze right now. In addition to robot-powered bars like Glacierfire popping up, we’re also seeing automated drink dispensing vending machines from Rotender and Celia start to hit the market. It’s not hard to see all of these types of robots working in tandem, however, with a robo-bartender pumping out standard cocktails, while Print a Drink prints up specialty drinks customized for special occasions. We’ll drink to that.

South Korea: Shinhwa Food Group to Use Woowa Bros. Dilly Plate Server Robots

The Shinhwa Food Group will be rolling out server robots at its restaurants brands in South Korea thanks to a new deal with Woowa Brothers. Aju Business Daily reported yesterday that Shinhwa has agreed to gradually introduce roughly 100 “Dilly Plate” server robots at its restaurants by the end of this year.

Dilly Plates are basically self-driving racks of trays that shuttle plates of food and empty dishes back and forth from a restaurant’s kitchen to its tables. The robots are equipped with sensors and smarts to do things like avoid obstacles and other people. Woowa debuted its first Dilly Plate (made by Bear Robotics, shown in the picture above) at a Seoul Pizza Hut back in 2018. According to the Aju Business Daily, Woowa now has five different Dilly Plate models that are deployed across 305 restaurants in South Korea. Woowa’s deal with Shinhwa is the biggest deployment agreement so far.

Interest in food robotics has accelerated since the pandemic last year. The use of a robot such as the Dilly Plate can reduce the amount of human-to-human contact in a restaurant by removing a human from the equation. Robots can also alleviate some of the hard work associated with being server by carrying heavy loads back and forth continuously without needing a break.

Woowa Brothers has been working on a number of different food robot-related initiatives. The company launched its food delivery robot program last year, and is even working with companies like Hyundai that allow those delivery robots to do things like autonomously access buildings and ride elevators. Additionally, Woowa worked with South Korean electronics giant LG and the Korea Institute for Robot Industry Advancement to develop its server bots.

A Woowa Brothers rep told Aju Business Daily that server robots are just the beginning of the company’s automation ambitions. Other plans include developing cooking robots and packaging robots that use artificial intelligence.

Boston University Researchers Develop Kirigami-Based Gripper for Robots

Researchers at Boston University (BU) have developed a new type of versatile robotic gripper inspired by the Japanese art of kirigami, or paper cutting. In a paper released earlier this month, Douglas Holmes, BU College of Engineering associate professor of mechanical engineering and BU PhD student Yi Yang demonstrated how their new type of flexible gripper can pick up a wide variety of objects including soft, perishable items like raspberries.

As shown in this video below, the new soft gripper is made by laser cutting a specific “shell” shapes out of flexible material. When the shell is placed over an object and both ends are pulled at the same time, the shell contracts, tightening around the object enough to pick it up, but not so hard that the object breaks. In the video you can see the gripper pick up a raspberry, a grain of sand and even rows of marbles.

This Robot's Soft Gripper Was Inspired By Japanese Kirigami

It’s not too hard to imagine different applications for this soft gripper technology in the food world. It could be used for harvesting berries, industrial packaging of food like eggs, or for picking and packing grocery items at an automated fulfillment center.

Food is often a good use case for robotics since it is oddly shaped and can be quite fragile. If a robot can manipulate berries or eggs or bread without breaking or squishing them, then that robo-dexterity can be transferred to other operations involving delicate materials.

There are actually a number of researchers and startups working to bring this level of precision to robotic grippers. Dexterity Robots can figure out how much pressure to apply to an object when picking it up to avoid smushing it. The aptly named Soft Robotics grippers use rubbery-tipped appendage and mimic an octopus to gently pick up objects. And a couple years back MIT researchers turned to origami, another traditional Japanese paper art form, to develop a cone-shaped gripper that acted similar to a Venus flytrap.

For those interested in learning more about Boston University’s kirigami-inspired robot, the full research paper can be found at Science Robotics (subscription required).

Kellogg’s Bowl Bot Cereal Robot, Based on Chowbotics Tech, Debuts at Two Universities

Kellogg’s Away From Home, a division of the CPG giant that works with foodservice operators, announced a joint pilot program with Chowbotics (now a division of DoorDash) that just debuted cereal dispensing robots at two universities. The aptly named Kellogg’s Bowl Bot will dispense a variety of Kellogg’s brand cereals as well as different milks, fruits, yogurts and more.

According to a press release sent to The Spoon, the Kellogg’s Bowl Bots are now available at Florida State University and the Univeristy of Wisconsin-Madison (which also uses Starship delivery robots!). The Kellogg’s Bowl Bot is basically a re-purposed and re-branded Chowbotics Sally robotic vending machine, which started out as a salad-making robot but has since expanded its capabilities to include other types of parfaits and bowl foods.

We’ve known about Chowbotics’ CPG ambitions since last year, when the company told us it wanted to partner with cereal companies specifically. At the time, the cereal bot use case was more about offering free samples at a grocery store, but branding and placing these machines at colleges makes a lot of sense. For one, colleges have long been a target location for Chowbotics. Prior to the pandemic, it had rolled out Sallys to a number of colleges and universities across the country. Second, the Kellogg’s Bowl Bot is perfect for our pandemic times since it doesn’t require human-to-human interaction and all of the ingredients are store inside closed containers that are themselves stored within the machine.

On top of all that, while my college days are long behind me, I assume cereal remains popular with this younger generation (when they aren’t eating avocado toast).

The Kellogg’s Bowl Bot has a special menu programmed for students that includes:

  • About Last Night with Kellogg’s Frosted Flakes, Kellogg’s Froot Loops, Kellogg’s KraveTM Chocolate, chocolate drops, banana chips, and espresso syrup
  • Hawaii 5-0 with Frosted Mini-Wheats, Bear Naked Fit Triple Berry Granola, pineapple, coconut, and mango
  • Valentine’s Day Vibes with Special K Red Berries Cereal, cocoa nibs, blueberries, strawberries, and whole milk

Or students can create their own custom bowl mixing and matching cereals, milks, nuts, seeds, fruits and more. Prices start at $2.99 and go up to $6.50, depending on selection. Students can even use their campus meal plans to purchase their bowl.

For those interested in checking out the new Kellogg’s Bowl Bots, the one at FSU is located in 1851, an on-campus dining facility and convenience store, and the robot at UW-Madison is situated on Dejope Residence Hall’s main floor, outside The Bean & Creamery.

Q1 2021: AppHarvest Bets on Robots, Strawberries and More Data in the Greenhouse

Control ag company AppHarvest is adding more of both crop types and technologies to its budding greenhouse network, according to the company’s Q1 2021 earnings call this week. That includes strawberries, leafy greens, harvesting bots, and lots of data.

The company, which went public in February, is best known at this point for the 60-acre greenhouse facility it operates in Morehead, Kentucky, where it grows beefsteak tomatoes. AppHarvest sent out its first shipment of these tomatoes to grocery stores earlier this year. Customers now include Kroger and Wendy’s.

CEO Jonathan Webb said on the company’s earnings call this week that two more Kentucky greenhouses, one in Richmond and one in Berea, will be operational next year, and that with them, AppHarvest will start growing leafy greens and strawberries. Webb pointed out that while his company may have started with tomatoes — a fairly traditional crop when it comes to greenhouse growing — the eventual aim is to “grow the company into a trusted high-tech sustainable food company.”

As far as that tech goes, AppHarvest’s CTO Josh Lessing said on the investor call that the company is investing in “robotics, artificial intelligence, teleoperation, and proprietary seed genetics.” To date, its biggest move has been the acquisition of Root AI, a startup best known for its crop-harvesting bot Virgo. (Lessing was the cofounder and CEO of Root AI before the acquisition.)

“Presently, we are training our intelligent robot Virgo to manage crops and inform growing decisions,” Lessing said on the call, adding that Virgo could eventually be configured to harvest multiple different crops, including delicate ones like strawberries — hence the company’s announcement to move into the realm of berry growing. 

As a crop, strawberries are highly suited to the controlled-environment agriculture (CEA) realm because they are extremely delicate, perishable, and normally require boatloads of pesticides when grown outside. Moving the grow process indoors, to a fully controlled environment, means better protection for crops from weather hazards, no pests and therefore no need for pesticides, and more consistent temperatures and humidity levels that can ensure better-tasting plants with a more robust nutritional profile. 

Given the amount of sunlight strawberries need for optimal growing, greenhouse settings are obvious candidates, since they rely largely on the sun with only supplemental LEDs. However, vertical farms, which use LEDs to mimic the sun’s light spectrum, are also now growing strawberries. Plenty, Oishii, and SinGrow are just a few of the names on that list. Whether one method will wind up superior to the other will (among other things) depend on what the end product tastes like as well as how much it costs to grow, sell, and buy.

For AppHarvest, though, the real win with technology will be not so much about the crops it can grow but the data Virgo and other tools can collect. That data can in turn get analyzed and turned into actions and insights applicable across the AppHarvest greenhouse network. “Granular plant level data from each fruit means we can learn exactly how to optimize quality, production, sales and logistics,” said Lessing. “This foundation will give us the opportunity to restructure the world’s food supply in order to mirror the hyper efficient e-commerce landscape.”

Along those lines, the company will expand beyond these first three facilities in the future. Two more projects will be announced this summer and are slated to be operational in late 2022. Webb said on this week’s call the company is on track to operate 12 greenhouses by 2025. By then, one imagines those facilities will grow a whole lot more than greens, strawberries, and tomatoes.