Close Ad

The Spoon

Daily news and analysis about the food tech revolution

Robotics, AI & Data

Kroger and Common Sense Robotics Each Announce New Grocery Robotic Fulfillment Centers

I get that it’s supposed to be three of something to make a trend, but the fact that two different companies a world apart made robot-fulfillment center opening announcements on the same day is totally indicative of a broader move towards grocery automation.

Here in the U.S., Kroger announced that Forest Park, GA, just outside of Atlanta, will be the next home of its Ocado-powered customer fulfillment center. Kroger will spend $55 million on this “shed,” as Kroger calls the centers, which will feature automated, robot-driven fulfillment of grocery orders. This is the third such shed of a planned 20 that the company plans to build. Other announced sheds are in Monroe, OH and Groveland, FL, with another one coming to the Mid-Atlantic region.

Over in Tel Aviv, Venture Beat reports that Common Sense Robotics has broken ground on a completely underground automated fulfillment center for an unnamed grocer. The new facility will be in a parking structure under the Shalom Meir Tower and will be 18,000 sq. feet. One of Common Sense’s selling points is that its vertically-oriented systems can better maximize available space and thus deliver full grocery store levels of product fulfillment in a fraction of the space.

That both of these stories happened on the same day is a coincidence, but it also highlights the moves grocery stores are making towards automation. Robotic fulfillment centers like these use totes on rails to quickly assemble items from online orders and hand them off to a human who puts them into bags for pickup or delivery. Robots can move faster than humans, they don’t get tired or need breaks, all of which can reduce the order fulfillment time down from hours to as little as a half hour for some systems.

This faster fulfillment is why so many grocery retailers are trying out robots. Takeoff Technologies has partnered with Ahold Delhaize and Albertsons, and Walmart is testing out automatic fulfillment through Alert Innovation. In each of those cases, robot-powered fulfillment centers are being built into the back of existing stores rather than off-site locations like Kroger and Common Sense’s.

These robotic fulfillment centers are very much in the early stages, but you can expect to see more of them over the next year as more grocers test and implement automation to get you your groceries faster.

Newsletter: In-House Vs. Third-Party Delivery App Showdown, The Rise of Personalized Menus

It’s only July 9 and I feel like I’ve already written “delivery is table stakes” this year so many times it’s lost some of its meaning. But it — the phrase and the delivery model itself — is here to stay, and businesses who aren’t at least working towards a strategy in that area will lose competitive advantage as more consumers migrate towards getting their food dropped at the front door.

That’s not a brand-new revelation. Restaurant owners, tech companies, and investors alike have been saying as much for some time. What is new, however, is that there are now more ways for restaurants to do delivery than the two extremes of pay for your own fleet or sign up with a third-party service. The new middle ground starting to get more attention is powered by technology.

A lot of it’s about using tech to drive more users to restaurants’ in-house apps, rather than to a third-party service like DoorDash or Grubhub, since restaurants risk losing some brand integrity and valuable customer data the minute they ink a deal with a third-party service. Companies like ShiftPixy, whom The Spoon talked to recently, are answering this problem by building software platforms that help restaurants drive more traffic and orders through their own mobile apps, but enable them to still take advantage of gig-economy-style driver fleets. ShiftPixy provides its own W-2 drivers to restaurants it works with.

Another company, Olo, offers a similar tech platform with a slightly different model, where they essentially partner with third-party services like Uber Eats and Postmates so that a restaurant doesn’t have to. The result could be the best of both worlds. Teriyaki Madness, for example, works with Olo and says it hopes to have 50 percent of its orders coming through its own mobile app and 50 percent coming through third-party apps by year’s end.

Third-party Delivery Apps Are Still Growing
Teriyaki Madness’ goal is a wise one to strive for right now, because even with increased talk of putting control back into the hands of restaurants, it’s clear third-party delivery services aren’t going away any time soon. That fact was underscored this week by eMarketer’s latest forecast, which predicts third-party food delivery apps will see upwards of 44 million users in the U.S. by 2020.

The forecast includes the usual suspects of third-party delivery and where they stand in the market, numbers that could swap around at any time, given how competitive food delivery is and how difficult attracting and retaining customers actually is, even for the top players.

What eMarketer does make clear, however, is that giving consumers more choices will be the key to driving growth for restaurants in the future.

Restaurant Menu Personalization Is on the Rise
An avenue for creating more choice in the restaurant will be through the menu itself. It’s a topic The Spoon’s Mike Wolf discusses often, and this past weekend he picked up that thread again on a podcast with Scott Sanchez, CEO of THE.FIT.

THE.FIT uses AI to help diners comb through restaurant menus and find items best suited to their dietary needs and preferences. While THE.FIT’s business is specifically focused on those with a lot of restrictions (e.g., Keto, gluten-free), this trend of using technology to personalize restaurant menus is one we’re going to see a lot more of. Large quick-service chains like McDonald’s are already employing AI-powered tech in some stores, while others, most recently Dunkin’, hint at similar initiatives for the future.

It will almost certainly play a role in the future of delivery. What that looks like, exactly, isn’t clear yet. But companies — whether a third-party service or a multi-unit chain using its own app — which use it to offer customers not just more choice but more relevant choice will be at a competitive advantage. At least, that is, until AI and personalization become table stakes themselves.

Until next time,
Jenn

Connected Robotics Raises ¥850M to Expand its Food Robot Lineup

Connected Robotics, the Tokyo-based startup that makes food robots, announced yesterday that it has raised a ¥850M ($7.8 million USD) Series A funding round. The round was led by Global Brain Corporation, with participation from 31VENTURES Global Innovation Fund, UTokyo Innovation Platform Co., Ltd., Sony Innovation Fund, and 500 Startups JP, L.L.C. This brings the total amount raised by Connected Robotics to ¥950M ($8.73 million USD).

Connected Robotics currently has two food robots: the OctoChef, which makes fried octopus balls known as Tokoyaki, a popular street food in Japan; and the Reita robot, which serves up soft serve ice cream. With the new money, Connected Robotics will accelerate research and development of new products: its automated dishwasher robot, a hot snacks robot for convenience stores, and an automatic breakfast cooking robot service dubbed “Loraine.”

The food service industry in Japan faces many of the same labor challenges as the U.S. Potential workers are avoiding the hard, repetitive, sometimes dangerous jobs at restaurants. Adding pressure to this labor crunch, Japan is facing an aging population, with 20 percent of its populace 65 years or older (and that number is projected to shoot up past 35 percent by 2050).

Robots and other automated systems can help alleviate this human labor shortage, and a number of companies are jumping in with their automated solutions. Sony has partnered with Carnegie-Mellon University to develop food robots. Miso Robotics’ Flippy is grilling and frying. LG is building a Flippy-like robot with CJ Foodville. And Dishcraft just unveiled its high-volume dishwashing robot.

We’ll actually be taking a first-hand look at food robots in Japan at our upcoming Smart Kitchen Summit: Japan next month. If you know of any good robots serving food in Tokyo, drop us a line and we’ll check it out.

Aryballe Raises €6.2 Million for its Digital Nose

Aryballe, the French “digital olfaction” startup that builds a device that essentially mimics the smelling power of the human nose, announced today that it has raised a €6.2 Million ($7M USD) Series B round of funding led by International Flavors & Fragrances (IFF) with participation from Hyundai Motor Company. This brings the total amount raised by Aryballe to €9.3 ($10.43M USD).

Aryballe debuted its NeOse Pro, a handheld device aimed at the B2B market that can be used to detect and identify odors, at CES in 2018. As Mike Wolf wrote last year:

The product works by attracting odor molecules into the device’s chamber where they then interact with chemical sensors. The device lights up the prism with an LED light and the device records optical signal transduction and then analyzes the odor signature and matches it against a database in the cloud.

In addition to the investment dollars, establishing the formal relationship with IFF’s massive database of fragrances will help augment Aryballe’s odor identification capabilities, and give IFF access to another application it can provide to its clients.

Applications for Aryballe’s technology include maintaining product consistency for something like coffee roasting — making sure each batch roasted smells the same, or ensuring authenticity of raw materials — or determining whether the vanilla a food maker received is artificial or natural.

Additionally, Aryballe is already talking with appliance manufacturers to place its digital olfaction into things like smart refrigerators and ovens. An embedded e-nose in these appliances could help detect food transformation, so your fridge would be able to smell when an item starts to spoil, or an oven could better know when something is done or burning. The company says we should start to see those Aryballe-enabled devices appear in 2020.

With the new money, Aryballe is looking to miniaturize its technology even further, perhaps embedding its technology into even smaller devices like food storage containers so they can give you a better sense of when your leftovers or fruits are going bad.

Aryballe is among a wave of companies looking to digitize our senses. Computer vision is used extensively in things like cashierless checkout in stores to see what we buy (and charge us automatically). Over this past weekend, IBM unveiled its Hypertaste platform to identify and classify liquids. Heck, Amazon has even applied for patents for its own refrigerator that smells.

While Aryballe is smaller than most of these other players, the relationship with IFF and its accompanying dataset, along with an actual working product, means it could lead the pack by more than an e-nose.

IBM Develops AI-Powered Hypertaste Tech to Identify Liquids

IBM announced over the holiday weekend that it has developed Hypertaste, an AI-assisted “e-tongue,” which can be used to identify liquids. This, according to IBM, can be helpful for situations where you want to test a substance… without actually putting in your mouth. Examples might be testing water quality at lakes or rivers, or identifying counterfeit wines.

Big Blue said that up until now, the issue with such testing was that portable sensors were too limited in what they could detect, and more powerful sensors were too big to be portable. Read the full blog post for a detailed explanation, but in a nutshell, Hypertaste works using a handheld sensor, a mobile phone app and the cloud. Put the Hypertaste device in a liquid and its electrochemical sensors (covered in polymer coatings) measure voltage in response to different combinations of molecules.

These recordings are sent to a mobile app, which shoots them up to the cloud where they are analyzed. Machine learning algorithms trained to know what that particular liquid should “look” like compare the sample, with results coming back in less than a minute. Hypertaste then can be used to test any liquid, as the algorithms just need to be trained on what to look for.

One of the applications for this technology is authenticating ingredients in food. From the IBM blog post:

…think of the supply chain safety from producer to consumer for packaged food and drinks. At present, once food and drinks are packaged, there is little ability to verify that the package actually contains what is on the label, apart from sending the product to a lab for testing. So, suppliers acting in bad faith may insert lower-quality products into the supply chain with little risk of getting caught, or counterfeiters may even fake a real product by adding the few analytes which are most likely to be tested for in a lab. Fooling a combinatorial sensing system such as Hypertaste is much harder as there is no single substance on which the identification relies, and it is more difficult for wrong-doers to access the sensor training parameters which provide the “key” to interpreting the chemical fingerprints.

E-tasting is becoming quite the cottage industry lately. The Chinese government has been using AI-powered robots to see, smell and taste food to ensure quality and authenticity. And in May, researchers at Washington State University developed an e-tongue for durable spiciness testing.

Right now, the applications for Hypertaste appear to be B2B based, but perhaps it could find its way into consumer products for home testing of water and other liquids.

Vegebot Can Identify and Harvest Lettuce (and More, at Some Point)

Homer Simpson once sang “You don’t win friends with salad,” but he might change his tune if he knew those greens were harvested by a robot.

Well, probably not. But we at The Spoon definitely think its cool that researchers at the University of Cambridge in England have developed a robot that can automatically identify and harvest heads of iceberg lettuce. Dubbed “Vegebot,” the research team announced over the weekend that its robot had successfully completed tests in a variety of field conditions on an actual farm.

Because it grows flat to the ground and is easily damaged, iceberg lettuce can be a challenging crop to harvest. The Vegebot uses an on-board camera and computer vision to identify the lettuce, whether it is mature enough to be harvested, and if it is healthy (diseased lettuce could spread to other lettuce in the harvest). Once identified, a second camera near the cutting blade makes sure the cut is smooth before a robot gripper picks it up.

This high-tech harvesting is cool, but as of now, it is also very slow. You can see just how slow in this video:

Robot uses machine learning to harvest lettuce

However, like with most things robotic, the speed will come. Right now, the robot has proven that its computer vision, cutting and gripping technology can work outside of the lab and under a variety of conditions. Additionally, the robot’s name is “Vegebot” not “Lettucebot” (there already was one of those), and Cambridge researchers say that the underlying technology could be used for a variety of crops.

Automation and robots are on the verge of becoming more mainstream in agriculture as technology like Vegebot’s, Augean Robotics and Agerris improve, and human labor shortages become more of an issue. Working on a farm requires intense physical labor, often in 100-plus degree weather. Having robots on hand can mean continued production without risking human life and health.

Even Homer Simpson can see that’s a winning idea.

FarmBot Launches Two New Robotics Kits to Give Anyone an (Automated) Green Thumb

The idea of growing my own fruits and vegetables is appealing, but the reality of that ever happening is pretty slim. I just don’t have the know-how or the patience to set up and manage a garden properly. But FarmBot, the startup that sells home gardening robot kits, could change all that as the company announced today the launch of its new Express and Express XL robots, and placed them on pre-sale for the month of July.

The FarmBot is a cartesian coordinate robot that sits on top of your raised garden. According to the company, each kit comes 95 percent pre-assembled, and should only take an hour to set up. Once that’s done, FarmBot’s software lets you drag-and-drop to map out where you want to plant your different vegetables and fruits. Then FarmBot will sow the seeds, monitor them through its on-board camera, water them properly, and even spot and remove weeds. As is par for the course these days, all of this activity can be tracked via mobile app on your phone.

The new FarmBot models won’t ship until November, but the company has put them on pre-order sale for the month of July. The Express (1.2 metes by 3 meters) will set you back $1,500 now, with the price jumping to $2,300 after the sale, and the Express XL (2.4 meters by 6 meters) will cost $2,000 during the sale, going up to $2,800 afterwards.

The other cool thing about FarmBot is that it’s completely open source — even its financials. All the CAD drawings, software and documentation is online for you to download. In fact, the entire company is open source and transparent about all of its actions and activities.

So far, FarmBot says that it’s shipped more than 1,000 kits to “early adopters” in more than 65 countries. FWIW, in December of 2018, the company reported that it sold a skosh over 800 of its previous Genesis and Genesis XL kits to bring in $2.5 million in revenue.

With its new Express and Express XL models, the company says it is “ready to bring FarmBot to the masses,” which includes not just residential customers, but small scale farmers and places like universities. And while $1,500 isn’t cheap per se, and the machine won’t be of any use for city dwellers because of its size, if the kits are as easy to set up and run as promised, that price point could get those aspiring — but failing — gardeners the robot green thumb they need.

FIBBEE is a Moscow-Based Robot Barista

If I told you that Russian ‘bots were getting more sophisticated and spreading, your first reaction might be one of concern. But what if I told you that those Russian ‘bots served coffee? Well, then you might welcome them with open arms.

Moscow-based Foodtronics has created FIBBEE, a robot barista that serves up all manner of lattes and other coffee drinks. Like Briggo and Cafe X here in the U.S., FIBBEE is an automated kiosk that can be set up in high-traffic areas, and like those American counterparts, the number of FIBBEE locations is expanding.

Moscow is downright hot for robo-baristas right now, as FIBBEE joins another automated coffee service, MontyCafe.

In our quest to chronicle the rise of food robots all over the world, following is an email interview with Foodtronics’ CEO Alexandr Khvastunov about FIBBEE. NOTE: Answers in this post were slightly revised after publication as Foodtronics felt the translations weren’t entirely accurate.

SPOON: What is Fibbee?
Aleksandr Khvastunov: FIBBEE is a robot-barista. We believe that it has a female entity so FIBBEE is her given name. Her personality is bright, easy-going and positive.

Besides the fact that she’s an ideal barista who never fails, FIBBEE may give you emotional recharge. She serves coffee to customer in one of the five colorful cups which are supposed to provoke different moods: boost, energy, balance, insight and fun.

How is Fibbee different from Cafe X, Briggo or even MontyCafe?
We’ve been professionally engaged in coffee for the past 8 years. We’ve opened traditional coffee houses (so-called third wave coffee houses), were engaged in the coffee wholesale and other projects.

At some point we understood that robotic retail may give us an opportunity to grow and scale well our love of coffee as of the lack of human factor – FIBBEE is fast, she’s always in a good mood and ready to meet your needs.

We have developed almost all technical components for FIBBEE, from chips and control systems to our own manipulators (specific system of moving objects).

We also have two coffee machine modules and we can brew two drinks at the same time. One manipulator inserts glasses into the coffee machine and when the drink is ready rearranges it into the waiting buffer. Another manipulator swaps coffee from the buffer to one of the 3 areas of issue.

How many different types of drinks can Fibbee make? Can it do hot and cold drinks?
Current menu includes espresso, americano, lungo, cortado, latte, cappuccino, cappuccino light, and coffee with syrups.

Still our customers can customize any drink. For example, they can add milk to black coffee in the required quantity, adjust the amount of foam and milk in cappuccino and latte, or select milk temperature (cold or hot).

In the nearest future we plan to introduce additional types of milk (soybean, oatmeal), various types of coffee beans and drinks with ice and nitro coffee which are being tested right now.

How many Fibbee’s are there and where are they located? How many more do you plan to roll out?
We have 2 robotic coffee bars now, and in July we’re opening the third. By the end of 2019, we plan to open more than 12 in Moscow. We’re creating our own chain and currently do not sell franchises.

Our ambition is to become the largest robotic chain of coffee houses in Russia and Europe. Because we love coffee and know how to make it delicious.

With Reported Tesco Partnership, is Trigo Vision Pulling Ahead of the Cashierless Checkout Pack?

Tesco, the U.K.’s largest grocery store chain, is reportedly testing out cashierless checkout using technology from Israeli startup, Trigo Vision. Bloomberg was first to report the as-yet-not-confirmed news, and framed its story as Tesco going up against Amazon and its Go stores. Which is accurate, but I think the bigger story, again if true, is about Trigo Vision gaining traction in the market.

Similar to the setup at Amazon Go stores, Tel-Aviv-based Trigo Vision installs high-tech cameras and artificial intelligence in existing grocery stores to automatically track what people purchase. Shoppers scan an app on their phone upon walking in, grab what they want and leave, getting charged automatically upon exit. Last August, Trigo Vision told The Spoon it had a pilot program with an unnamed European store, and was aiming to to open a full store in Europe by August of this year. Knowing that, word about a quiet partnership with Tesco getting leaked to the press now would make sense.

If the Bloomberg story is true, this would be the second major grocery chain partnership for Trigo Vision. In November of last year, Shufersal, Israel’s largest grocery chain, announced it would be rolling out Trigo Vision’s tech to its 272 store locations over this year. You can actually see Trigo Vision in action at a Shufersal in this video from an Israeli news broadcast:

As we’ve covered before, there are a raft of startups all vying to help supermarkets move to cashierless checkout. But Trigo Vision is the only one to have named at least one partner publicly. Others, such as Grabango, Caper, AWM Smart Shelf, Zippin and Standard Cognition are all in tests with retailers, but those retailers are still unnamed.

Again, the Bloomberg story still needs to be confirmed (we’ve reached out to Trigo Vision for comment), but perception can play a big part in vendor selection. As Trung Nguyen, VP of eCommerce for Albertsons explained at our recent Articulate food automation summit, grocery retailers aren’t just looking for cool new tech; they need a solution that can work at scale. If Trigo Vision’s name keeps coming up in headlines as a proven technology partner, then Trigo Vision will become the perceived leader of the cashierless checkout pack. This perception can then lead to more deals and that dominance can perpetuate itself.

Now, it’s still very early days in the cashierless checkout space, and Trigo Vision’s long-term success over other vendors is far from guaranteed. As noted, there are a lot of players with different takes on the tech and there are a lot of grocery stores in the world. And while Amazon and Walmart are developing their own cashierless checkout solutions, the sheer scale of their overall businesses could be a threat to the very existence of other retailers.

Tesco, it seems though, is ready to put up a fight.

Mmuze Expands its AI-Powered Voice Shopping Platform to Groceries

Though we have a number of Amazon and Google voice assistants scattered throughout my house, the only shopping we’ve ever done is when my 8 year old son once ordered hot chocolate via Alexa, mostly because he thought it was neat. And my family isn’t a fringe case, voice shopping has been slow to take off with surveys showing people are concerned about privacy, order accuracy, or are just not interested in buying stuff by speaking (raises hand).

Mmuze (pronounced “M-uze”) is looking to change all that with its conversational AI technology that allows retailers to offer shopping by voice or text. The company has been working with retailers like Jordache and Perry Ellis and today announced that it is expanding into grocery. Through Mmuze’s APIs, grocery stores can offer voice/text shopping through their own mobile app, web site or smart speaker app/skill.

“The problem with most voice technologies happening today,” Ran Zfoni, CEO and co-founder of Mmuze told me by phone, “is they are not designed for retail, so they are generic. They intend to solve many different cases.”

To challenge this generic approach, Zfoni told me that the first thing Mmuze does is create a domain expertise around a particular vertical. “A shopping dialogue for pasta is not the same as ordering an Uber,” Zfoni said.

There are two parts to obtaining this domain expertise. First is the basic level of understanding a grocer’s catalog of products. What they are, how much they cost, different brands, etc. The second layer is bringing in data from the social web to understand how people are using these products to better understand a shopper’s intent.

For example, a shopper could be ordering hot dogs, hamburgers and buns, so Mmuze’s chatbot will ask if they are hosting a barbeque because it understands that those items are often used at cookouts. If they are, Mmuze can then suggest other items such as charcoal, or it can access your purchase history to see if you want to re-order particular items.

Additionally, Mmuze can also act as a regular shopping list. The technology can work across platforms so you could add to your list of groceries to get throughout the day as you remember or run out, no matter what platform you are using. You can then complete your purchase and have the groceries ready for pick up or delivery.

AI-powered recommendations is a hot topic right now. Halla is another startup using machine learning to make better recommendations for grocery retail apps. While it’s a different type of retail, over in the fast food world, Clinc raised $52 million last month for its conversational AI for drive-throughs, and earlier this year McDonald’s bought Dynamic Yield to have their menus make smarter recommendations.

Founded in 2014, the Tel-Aviv-based Mmuze started as part of a Microsoft accelerator and was also part of the eBay Innovation Lab. The company has raised $4 million in funding. According to Zfoni, Mmuze is focused on the U.S. and Europe and is working with unnamed grocery retailers in tests. Mmuze will have a usage-based pricing model, but didn’t disclose any specifics at this time.

If Mmuze’s technology works as promised, perhaps it could give voice shopping a shot in the arm. But the biggest barrier to getting grocers to adopt the technology could be the grocers themselves. As Trung Nguyen, VP of eCommerce for Albertsons said at our recent Articulate food automation summit, the retailers don’t just want great technology, they also want easy implementation. If Mmuze can get that piece as well, then they’ll definitely have something to talk about.

YPC Wants to Bring Fast Food Robotics to Fresh Food Cooking

We know that robots can cook up hamburgers, fry tater tots, and even make delicious bowls of food in high volume restaruants. But YPC Technologies wants to put that fast food style of robot to work making more complicated dishes like salmon filets, mushroom risotto, or even raspberry sorbet.

Based in Montreal, Canada, YPC (which stands for Your Personal Chef) has built a robotic workstation that its says can make high quality complex, fresh-cooked meals. Right now the YPC robot uses an articulating arm which grabs ingredients, pours them out into various multi-cookers and other devices that do the chopping, stirring and cooking. The YPC robot can make thousands of recipes and, depending on the complexity of the dishes ordered, can cook roughly 100 dishes per hour.

For all of its robotic bells and whistles, however, YPC Co-Founder and CEO Gunnar Grass told me by phone that YPC is not intended to be a fully autonomous kitchen. Humans will still be around for tasks like re-stocking ingredients and doing the final presentations. “The plating of the dishes is very difficult to achieve with robots,” Grass said, “in the long run we’ll automate 60 percent of kitchen operations.”

Grass stressed that this version of the cooking robot with the articulating arm is very much in the prototype phase, and will go through much more innovation, including the addition of a two-axis arm. Eventually, Grass said YPC will probably take up 40 sq. meters (a little more 400 sq. ft.), so it’s around the same size as the PAZZI automated pizza restaurant.

At first, YPC wants to own and operate its own robot eateries and is targeting mid-volume traffic areas like co-working spaces and retirement homes. It might also partner with a food service operation like Sedexo to be in university food halls that have more than one dining option. The YPC is not meant for high-volume sites like arenas or cafeterias that service thousands of people at once.

Co-working spaces actually seem like an ideal environment for a YPC system. There are plenty of office workers who want a good meal without having to leave the office, but there aren’t so many orders as to overwhelm the articulating arm.

While robots like Flippy are already working shifts, and Creator and Spyce are robot restaurants already open to the public, YPC is still very much in its early stages. Grass said his company raised a pre-seed round of funding, and that the operating prototype still needs to get the appropriate regulatory licenses before they go more public.

YPC is illustrative of the fact that automation in dining won’t be limited to just high throughput venues like fast food restaurants and arenas. Eateries that service smaller, but steady customer bases will also be able to take advantage of food robots, and will be able to provide a wider variety of meals than just burgers and tater tots.

McDonald’s is Testing Kitchen Robots and AI-Powered Drive-Thrus. It’s About Time

Last week the Wall Street Journal broke the story that McDonald’s is testing robotic deep fryers and voice-activated drive-thrus at a location in suburban Chicago.

The story didn’t lay out a ton of details. We don’t know what the robotic fryer will look like — whether it’s akin to Miso Robotic’s Flippy, which fries tater tots with an articulating arm, or if it’s more of an automated basket that dips and raises from hot oil. Likewise, all we know about the new drive-thru tech is that it’s a type of voice recognition technology.

But even with those question marks, this is a huge step forward for automation in fast-food. It’s also an inevitable one, since the success of the QSR is predicated on making consistent food at a fast pace. I’m hard-pressed to find a task that robots are more suited for.

The WSJ pointed out that these innovations are part of McDonald’s efforts to quicken the chain’s pace of service to help it edge out fast-food competition. That’s certainly true, but there are numerous other benefits as well. Automating dangerous jobs like frying could lead to fewer employee injuries. In addition to being potentially life-threatening, these injuries can also have a high cost for the restaurant.

As far as the drive-thru goes, adding voice recognition technology can help McDonald’s streamline the drive-thru experience, cutting down on ever-lengthening wait times. It could also pair up with Dynamic Yield, the AI-powered personalization platform that McDonald’s acquired a few months ago, to access customer’s past orders, preferences, and dietary restrictions to better upsell them on targeted add-ons.

Of course there will still be kinks to work out. Voice technology isn’t perfect, and robot fryers still need human employees to help them do things like bag orders. McDonald’s is also already getting blowback from skeptics who worry that automation will take valuable jobs away from humans.

However, speakers at our ArticulATE conference this April pointed out that there’s a growing labor crisis in the restaurant industry. QSR’s especially are struggling to find people to do the more repetitive, boring and dangerous tasks — like frying nuggets and taking orders for hours at a time.

McDonald’s isn’t the first restaurant to try out cooking robots or voice ordering technology. However, it’s by far the largest. It’s hard to argue against the fact that automation in food-service, as with any industry, was always going to happen eventually. But once a giant like McDonald’s starts putting the automation wheels in motion, eventually starts to look a whole lot sooner.