Close Ad

The Spoon

Daily news and analysis about the food tech revolution

Robotics

The Food Tech Show: Are We Ready to Eat Bugs?

The Spoon team got together talk about the most interesting food and kitchen tech stories of the week, including:

As always, you can listen to the Food Tech Show on Apple Podcasts, Spotify, download direct to your device or just click play below.

Drive-Thrus Are Getting Slower. Can Tech Change That?

Wait times in the drive-thru lines are getting slower, according to QSR’s annual Drive-Thru Study, which launched this week and says drive-thru speed of service in 2019 was 20 seconds longer than in 2018.

The study actually covers a number of different areas of drive-thru performance, from customer service to order accuracy to which chains are installing digital menu boards. But the continued lag in speed of service stuck out this year.

In 2019, the average time a customer spent in the drive-thru — from speaker to order window — was 255 seconds. To put that number in context, the average time from speaker to order window in 1999 was roughly 181 seconds, according to data from previous QSR Drive-Thru studies. That number dipped up and down over the next decade before climbing to 226.30 seconds in 2016. It’s gone steadily upward ever since.

What’s with the wait?

One reason is menus. The QSR 2019 study notes that “more complex menus” contributed to slowdown in order accuracy, which fell just over 5 percentage points compared to 2018. As the study says, “more intricate menus touted by brands like Taco Bell and Arby’s proved to be a stiffer challenge for employees working to deliver complicated orders at top drive-thru speed.”

Along with those larger menus come more complicated food items, like Taco Bell’s now-retired XXL Grilled Stuft Burrito, the making of which adds more time to the drive-thru process. Even a latte, which isn’t an inherently complicated drink, takes more seconds to make than pouring a regular cup of coffee. Some chains brands have gotten hip to this problem and trimmed down their bloated menus. Even so, it ain’t 1985, when you could count McDonald’s burger offerings on one hand, and we’re not likely to return to that level of simplicity.

Mobile orders also contributed to the slowdown in drive-thru times, and with “lanes possibly getting more crowded with not only drive-thru customers but also those picking up mobile orders, it’s going to be difficult for brands to shave off seconds moving forward.”

Some are trying to knock out those extra seconds, most notably Dunkin’, which started building stores with dedicated drive-thru lanes for mobile-order customers in 2018. Fellow donut-peddler Krispy Kreme is doing the same as it revamps its locations for the digital age. In Australia, KFC is piloting a drive-thru-only store heavily focused on digital transactions and testing out new concepts to speed up wait times.

But mobile orders aren’t going away, and, as I mentioned earlier, menus aren’t about to get smaller, so what’s a restaurant chain to do?

Implement AI.

McDonald’s made that much clear when it acquired Dynamic Yield in March and subsequently rolled out the latter’s AI tech to hundreds of locations. Right now, that particular implementation of the tech is aimed at things like order accuracy and quickly upselling items to customers. Where it could make a massive impact, though, is in making restaurants more predictive.

Dynamic Yield-enabled menu displays at McDonald’s drive-thrus can show customers food based on data like the time of day, the weather, and trending menu items. Narrowing down someone’s food selection based on those factors could help the average customer parse through a massive menu and get through the selection process faster, shaving seconds off the time between ordering and collecting the meal. As McDonald’s CEO Steve Easterbrook noted on an earnings call in April, “. . . using the data collected based on current restaurant traffic at the drive-thru, the technology will begin to suggest items that can make peak times easier on our restaurant operations and crew.”

It could also help predict future demand. If the system can tell by data that it’s 80 degrees outside, sunny, and a football game is about to let out nearby, the restaurant can prepare itself with extra staffing ahead of time to move a potentially bigger rush through the line faster.

Meanwhile, companies like 5Thru and Valyant AI are implementing things like license plate recognition and conversational AI to speed up the order and pay process in the drive thru. 5Thru, in particular, is also working with car manufacturers to add voice-order capability in the vehicle, sort of like Domino’s is doing with Chevrolet and other car companies.

All of these efforts are aimed at automating parts, or eventually all, of the drive-thru process to keep chains competitive in what’s become a very oversaturated fast food market. This time next year, we’ll have a better idea of how far towards that goal Dynamic Yield can get McDonald’s, and AI in general can get the industry. If the seconds spend in line start to drop, we could one day offer a 2020-sized menu much faster, so restaurant operators can party like it’s 1999.

Tally ho! Simbe Robotics Raises $26M for its Inventory-bot

Simbe Robotics, which makes the autonomous Tally inventory robot for retailers, announced today that it has raised a $26 million Series A round of funding led by Venrock with participation from Future Shape, Valo Ventures, and Activant Capital. Additionally, Simbe also announced today that it is expanding its existing partnership with SoftBank Robotics America to include inventory financing to scale the manufacturing of an additional 1,000 Tally robots over the next two years.

Tally is an autonomous robot that roams store aisles using computer vision and RFID to scan shelves to check on inventory. It’s part of a suite of services from Simbe that provides analytics about purchases and insights about re-stocking management. Simbe provides the hardware for free and charges a monthly subscription for the software and analytics. So far, Tally has been put to work in trials at Giant Eagle and Schnuck’s grocery store chains.

Tally is just one of the robots coming to a grocer near you. Earlier this year, Walmart announced it would expand Bossa Nova’s shelf-scanning robots to 300 locations, and Ahold Delhaize ordered 500 “Marty” floor roaming robots from Badger Technologies.

While retailers may like the fact that robots are faster and more precise than humans (and the fact that they don’t take breaks or call in sick), there are still a lot of kinks that need to be worked out with robots. As we’ve written before, robots can make their human co-workers enjoy their jobs less, and shoppers don’t know how to interact with a cold, silent, sentinel (even if they have googley eyes).

Additionally, how long will robots be necessary for things like inventory management? Walmart debuted its IRL store earlier this summer which features banks of cameras installed and computer vision to keep a real-time eye on what’s in stock.

However, retrofitting a store with the cameras, software and sensors needed to keep track of inventory in that way is expensive and will take a long time. So until then, robots like Tally will probably find a place among the produce at plenty of retailers.

Editor Roundtable Podcast: We Have Opinions on Tiny Dishwashers & Beyond Burgers

In case you didn’t already know, The Spoon team has lots of opinions. As you might guess, those opinions are especially pronounced when it comes to food and kitchen gadgets.

And so we decided to get together and get some things off our chest on this Editor Roundtable edition of the Food Tech Show.

Here’s what we talked about:

In addition to lots of opinions, we also share sound effects (or at least I do).

As always, please subscribe, play (and rate!) the podcast in Apple podcasts (or your favorite pod player), download direct or just click play below.

Video: Do Food Robot Startup Founders Need Restaurant Experience to Be Successful?

Say you’re a VC looking to invest in a company that makes strawberry-picking robots. There are three or four companies in the space, all vying for your capital to get off the ground.

How do you choose where to put your dollars?

That’s one of the questions that we tackled last week at ArticulATE, our inaugural food and automation summit. We closed out the day with a lively panel on the opportunities — and challenges — in the food robotics investment landscape. Our speakers were VCs from the foodtech, hard tech and IoT spaces: Brian Frank of FTW Ventures, Brita Rosenheim of Better Food Ventures, Rajat Bhageria of Prototype Capital, and Avian Ross of Root VC.

The Spoon’s Michael Wolf moderated the conversation on what these investors are looking for in a food automation startup pitch specifically — and where they see significant opportunities in the fast-growing market. (Yes, both Frank and Ross have invested in strawberry-picking robots — and Ross claims he made the right choice.)

It’s certainly an exciting time in the food robotics space: there are tons of entrepreneurs out there with lofty plans to build the next robotic sushi restaurant, the next automated food delivery bot, or the next burrito-rolling robot arm (which, apparently, really hard to do).

However, the panelists seemed to agree that food robotics is a trickier investment space than a lot of other tech areas. Sure, the basic building blocks of food robotics — AI, articulating arms, etc. — are pretty democratized. But Ross (who — fun fact — spent a former life building robots for the Food Network) said that “robotics feels special and different.” He pointed out that the food system is incredibly complex and that a whole host of players have to be involved to deliver even the most basic meal to the consumer. And that’s just logistics: getting a robotic system to reach parity with a basic human fast food experience in terms of taste or customer experience is really tricky.

Because there are so many complexities at play it can require more capital than some other tech investments. It can also take longer to bring food automation technology to market. Which isn’t a problem — unless, as Rosenheim pointed out, you’re working with investors who are looking for “the next shiny thing” and aren’t patient enough to be in it for the long haul.

Investors in food robotics have to be especially willing to take risks and play the long game. However, not all the VCs saw eye-to-eye on what it takes for a food automation startup to be successful. The panelists disagreed on whether or not startups need deep restaurant market knowledge to be successful, how high the capital investment has to be in food automation, and what sets one seemingly identical food robotics startup apart from another.

Check out the video below to see the whole conversation — it was a really fun one.

Articulate 2019: Investment Opportunities in Food Robotics

Look out for more ArticulATE 2019 videos rolling out on our YouTube channel over the next week! 

Kroger and Ocado Begin to Roll Out Automated Fulfillment Centers

If you’re as fascinated as we are with the idea of robots scurrying around and doing your weekly grocery shop, Kroger and Ocado have some good news for you.

Today the grocery giant announced it would roll out two new Ocado-powered customer fulfillment centers (CFCs) — also called “sheds” — in the Central Florida and Mid-Atlantic regions. This news comes just a few months after Kroger named Monroe, OH (outside of Cincinnati) the location for its first automated robot warehouse.

In May of 2018 Kroger entered into an exclusive partnership with U.K.-based online grocer Ocado, nabbing a 6 percent stake in the company and promising to build twenty automated warehouse facilities across the U.S. over the following three years. We’re seeing that promise come to life.

As resident robo-expert Chris Albrecht pointed out, grocery logistics is hot hot hot right now and big players like Walmart, Target, Amazon, and Kroger are all racing to deliver your groceries as fast as possible.

To beat out its competitors, Kroger has invested not just in robotic fulfillment centers but also in self-driving cars, a direct-to-consumer e-commerce platform, and an expanded deal with Instacart.

And that’s just for grocery delivery and curbside pickup. In-store, Kroger also has a number of developments in the works. In this month alone, they partnered with Microsoft to launch two tech-enabled grocery stores (their answer to Amazon Go), and also unveiled new Customize It personalized meal kits. With the company’s brand new innovation lab, which it launched in August in tandem with the University of Cincinnati, Kroger doesn’t seem like it will stop reinventing the grocery space anytime soon.

Curious about how robotics and automation will reshape the grocery business? Join the conversation at ArticulATE, our food robotics summit in San Francisco on April 16th! Early Bird tickets are available now.

CES 2019 Video: Watch the Rotimatic Make Tortillas

We were excited earlier this week when Zimplistic announced that its Rotimatic flatbread maker would be adding flour tortilla recipes to its repertoire. We were even more excited when the company brought the Rotimatic to our Food Tech Live event so we could try them firsthand.

They did not disappoint. Using just flour, oil and water, the Rotimatic can mix, knead and bake fresh tortillas in just 90 seconds. The result is a warm, fresh-tasting tortilla. They come out a wee bit crispy, so they aren’t like the kind you get at a taqueria, but they would be a nice option for your next taco Tuesday (or any occasion, really).

The flour tortilla recipe hasn’t been released yet into the wild, so existing Rotimatic users can’t make them today, but the update should be arriving soon. Now the question is whether the addition of flour tortillas will entice new buyers to pony up a thousand bucks for the device.

Watch the Rotimatic Make Flour Tortillas

The BreadBot Is A Bread Factory for the Corner Grocery Store

Back in college, whenever I rode my bike through the Gasworks Park neighborhood in Seattle, I’d get hit with the smell of bread baking as I passed the local bread factory.

As you can imagine, it was wonderful.

And now, Wilkinson Baking Company hopes to bring the best smell in the world to your local grocery store with its new bread making robot, the Breadbot.

The BreadBot, which made its debut last night at CES Unveiled, is a standalone bread robot that can make up to ten loaves of fresh bread per hour from scratch. The machine is fully automated, so after the store employee adds in bread mix to the mix hopper, the BreadBot mixes, forms, proofs, bakes and cools the loaves of bread.

The bread robot was designed to make bread pretty much anytime during the day (or night), and could be scheduled to start the process up to three hours before a store opens. So when the early rising morning shift worker unlocks a store with a Breadbot inside, not only will the smell of fresh bread waft out, but over a dozen loaves will be waiting to be sold.

“It’s the first time bread can be made in a fully automated fashion, start to finish,” said Dr. Randall Wilkinson, the CEO of Wilkinson Baking Company.

Wilkinson envision the BreadBot going into grocery stores, where the machine can sit in the bread aisle and make bread in front of customers. The company’s been trialing the breadbot in local grocery stores in eastern Washington and the results are basically what you’d expect.

“You see the crowds here at CES, that’s what happens in the store,” said Wilkinson. “The consumer comes in, says ‘what’s this?’, and if you’re a retailer you say, ‘that’s what I want.'”

The BreadBot is part of a larger trend of automation and robotics enabling more localized creation of food types that are normally made in a big factory somewhere. Much like the Bellwether coffee roasting machine can push coffee roasting out to the local neighborhood coffee shop, the Breadbot allows the local grocery to offer fresh made bread made in-store.

“Imagine taking a $150 million bread factory, slicing it up into small pieces, move that out into the grocery store, and make it fresh there rather than in the factory,” said Wilkinson.

You can see my full interview with Dr. Wilkinson and see shots of the BreadBot in action in the video below (bread smell not included):

BreadBot, a bread making robot, debuts at CES

Spoon Newsletter: LG’s HomeBrew Appliance, Spinn Update, World’s First Cell-Grown Steak

This is the post version of our weekly newsletter. If you’d like to get the weekly Spoon in your inbox, you can subscribe here.

When it comes to smart kitchen startups, there is no shortage of companies trying to change how we make our morning cup of joe.

One of the highest profile startups in this space over the past couple of years is Spinn, maker of a grind-brew coffee machine that utilizes a patented centrifugal brewing technology. I was intrigued enough with the company and its tech to plop down over $300 to join the first “batch” of orders in 2016, but almost two years past the promised ship date, I’m still waiting for my Spinn.

The company has hit some snags as they work to get their coffee maker to customers. While many of the Spinn’s early customers have been understandably frustrated, it appears most are hanging on, intrigued by the company’s periodic updates showing progress on the product. If they’re like me, I suspect many are getting frustrated with a company that continues to advertise and sell their overdue appliance, all the while creating more and more distance between new customers and those still waiting at the front of the line.

Still, I know I have no one to blame for investing my money in a Spinn with the knowledge that coffee startup products have been historically risky investments. With companies like ZPM and Arist turning out to be colossal misses and others like Bonaverde sputtering along for years and periodically shipping out products, I knew full well that I might never see the product. In the end, I may have been better off putting my money towards a Tesla like Spoon reader Rebecca:

“I put money down for this machine in Nov. 2016 (I’m a 1st batch-er) and at around the same time my husband put down money for a Tesla Model 3. At least I now have a smooth ride to the coffee shop!”

You can read my latest update on my Spinn journey here.

Coffee isn’t the only category that can be rough sailing for new startups. The home brew market has been notoriously tough as well, with companies like HOPii and iGulu struggling to ship and others like BrewArt and Brewie failing to get much traction.

Still, this hasn’t stopped South Korean consumer electronics giant LG from jumping into the game. The company recently announced they would debut a new home brewing device by the name of the LG HomeBrew at CES.  The capsule-based beer brewing appliance, which makes roughly the same amount of beer as a PicoBrew Pico per cycle, is expected to ship sometime next year.

What’s interesting to me is a company like LG usually only gets into a business if they see a high volume opportunity, which begs the question what will they do to differentiate their product?  While PicoBrew has certainly raised awareness around automated beer brewing, the company has yet to make home brewing a mass-market hobby, something I am sure LG hopes they can do.

Either way, next year promises to be an interesting one when it comes to tech-powered boozing at home. About a month ago, Keurig and AB InBev launched Drinkworks, a capsule-based instant serve cocktail, cider and beerbot machine expected to ship in limited quantities next year.

While home bev-tech space continues to slowly gestate, the pro market continues to move along at a rapid clip. Just this week, a robot powered bartender by the name of UR5e debuted at Broncos Stadium. The new bartenderbot is essentially robotic arm that grabs a beer cup and fills it (through the bottom, no less) for waiting customers.

Moving beyond beverage bots, this week also saw the introduction of a new sidewalk delivery bot from Postmates that goes by the name of Serve. Serve, which looks like the lovechild of Minion and Starship deliverybot, is expected to roll out in Los Angeles over the next year. And speaking of food delivery, this week Chris wrote about the breakup of Amazon and Instacart, something that’s been rumored ever since the big online retailer acquired Whole Foods.

There’s lots more great analysis to catch up from this past week, including looks at the first cell-grown steak and what the new farm bill means for CBD, so make sure to check it out.

That’s it for now. Have a great weekend and we’ll see you next week!

Mike

In the 12/14/2018 edition:

Presto Eats May Be the Most On-Trend Meal Kit Company Yet. But Will It Succeed?

By Catherine Lamb on Dec 14, 2018 11:16 am
Whenever I get word about a new meal kit company, it’s hard not to be immediately skeptical. It’s no secret that meal kits are struggling: Chef’d surprised everyone when it shut down abruptly earlier this year. Boston hyper-local meal kit Just Add Cooking ceased operations this fall. And Blue Apron’s stock continues to underwhelm.

The Denver Broncos Get a Beer Pouring Robot at Mile High Stadium

By Chris Albrecht on Dec 14, 2018 08:04 am
While the Denver Broncos may be in the midst of a losing season, they could win over fans this weekend when a new robot starts dispensing Bud Light at Mile High Stadium (h/t The Washington Post). You’d think that such a mechanical miracle would have a fancy name like the “Robo-Bronco” or the “Elway 3000,” […]

Startups! They’re Just Like Us! Amazon and Instacart Break Up

By Chris Albrecht on Dec 13, 2018 04:00 pm
Like Ben and Jennifer, Brad and Angelina, and Cardi B and Offset*, Instacart and Amazon have broken up. In a blog post today, Instacart announced that it was winding down (consciously uncoupling?) its grocery delivery relationship with Amazon. Like with so many other power couples, this breakup wasn’t entirely a surprise.

The Farm Bill Just Passed — What Does That Mean for the CBD Market?

By Catherine Lamb on Dec 13, 2018 02:08 pm
After months of back and forth, Congress voted yesterday to pass the 2018 Farm Bill. The $867 million bill contains lots of wide-reaching legislative measures, like expanded farm subsidies, SNAP revisions, and permanent funding for farmers markets. But perhaps most interestingly, the bill legalizes the production and sale of hemp at a federal level.

What Bowery’s Latest Funding Round Says About Indoor Farming

By Jennifer Marston on Dec 13, 2018 12:00 pm
New Jersey-based indoor-farming startup Bowery announced yesterday that it has raised $90 million in fresh funding. The round was led by Alphabet Inc.’s GV with participation from Temasek and Almanac Ventures, General Catalyst and GGV Capital (Bowery’s Series A investors), and various seed investors. Bowery produces what founder Irving Fain calls “post-organic produce.”

ImpactVision Raises $1.3M Led by Maersk

By Chris Albrecht on Dec 13, 2018 10:11 am
ImpactVision, a startup that uses hyperspectral imaging to assess food quality, has raised $1.3 million, according to VentureBeat. The round was led by logistics and transportation company Maersk, and brings the total amount raised by ImpactVision to $2.9 million. As we wrote last year about ImpactVision: Using a combination of digital imaging, spectroscopy and machine learning, […]

Postmates Debuts its own Bright-Eyed Delivery Robot

By Chris Albrecht on Dec 13, 2018 06:00 am
You know what you can look forward to in 2019? More robots (but more on that in a later post). Case in point: delivery service Postmates announced today that it has developed its own autonomous delivery rover that will be hitting sidewalks next year. Dubbed Serve, the li’l robot is a bright yellow square-shaped box […]

Hi Fidelity Genetics Raises $8.5M for AI-Driven Plant Breeding

By Chris Albrecht on Dec 13, 2018 04:00 am
Hi Fidelity Genetics (HFG), which combines sensors, data and artificial intelligence (AI) to improve plant breeding, today announced that it has raised an $8.5 million Series A led by Fall Line Capital and Finistere Ventures. This brings the total amount raised by HFG to $11.5 million. There are two parts to the HFG system.

Deliveroo Goes Back to the Future with Brick and Mortar Food Hall

By Catherine Lamb on Dec 12, 2018 04:00 pm
You know how they say that in fashion, everything old comes back around and eventually is new again? It seems that the same might be true for restaurants. This week London-based food delivery startup Deliveroo opened up its first brick-and-mortar location in Hong Kong (h/t CNBC). The so-called Delivery Food Hall is home to five […]

Waitr to Acquire Bite Squad for $321.3M

By Jennifer Marston on Dec 12, 2018 02:00 pm
Online delivery platform Waitr has announced plans to acquire third-party delivery service Bite Squad for $321.3 million. The purchase price is a mix of cash and shares of common stock of Waitr. Both companies serve small- to mid-sized U.S. markets. Waitr is a full platform from online order and delivery. Restaurants partnering with the company […]

Robots, Instagram, and Flan: Highlights from our Future of Restaurants Meetup

For our latest food tech meetup, we decided to do things a little bit differently. We were lucky enough to work with Chef Eric Rivera (formerly of Alinea in Chicago) and host the event at his new incubator space, addo.

We knew Chef and his team had high standards, but we were still blown away when we walked in the door. Addo is a creative culinary mind’s dream: the space is a coffeeshop during the day and plays host to pop-up dinners, cooking classes, and community get-togethers (Mario Kart tournaments, salsa classes, etc.) during the evening and weekends. “Basically, it allows me to do whatever I want,” he said.

That means that he lets chefs — usually ones who are up-and-coming or have been recently let go — to host pop-up dinners there, and fills the pastry case with locals who want somewhere to display their baking chops. The space opened its doors 2 months ago and already has a staff of 20.

Part of the reason that Rivera could even create a place like addo is because of new technology. When he returned to Seattle to cook after working in the legendary Alinea, where he was Director of Culinary, Rivera hit a lot of roadblocks. People told him he couldn’t start a restaurant without a certain amount of money, or investors. But Rivera decided to forge ahead and create a space inspired by his own struggles — powered, at least partially, by technology. Culinary booking services like Tock allowed him to do things like host 2-seat dinners out of his home kitchen before he got the addo space, and food delivery from his current operation pads his business.

A schedule of addo events
The pastry case

Over a Puerto Rican meal of roast pork with chimichurri, fried plantains, and flan (Rivera’s mom’s recipe), the chef joined The Spoon’s Michael Wolf and Modernist Cuisine’s Scott Heimendinger in a conversation about how technology big and small is changing the restaurant — from robotics to sous vide to Instagram.

Instagram may seem relatively banal when it comes to restaurant tech — after all, most of us use it to post food pictures all the time — but for bootstrapped entrepreneurs like Rivera, it’s been game changing. “I don’t have the money to hire a marketing or PR firm,” said Rivera. Instead, he uses Instagram as a marketing platform, as well as a way to target certain demographics.

It’s also a tool for him to tap into another trend we cover a lot in the food tech space: personalization. When diners reserve a spot online, Rivera has them fill out a questionnaire to get their dining preferences, but he also does a little sleuthing on their social profiles to see what restaurants they like to eat in, food preferences, etc. It’s a little Minority Report-y maybe, but when you’re paying a hefty price, you want each dish on the menu to be something you really want to eat.

Chef Rivera in front of his beloved smart oven.

Of course, we also had to cover a contentious subject for the future of restaurants: robots. “Robots are better than people for almost everything,” said Heimendiner. Especially for when you need to do something that’s highly repeatable or requires lots of accuracy.

Which is obviously useful in places with high volume and basic tasks, like flipping burgers. In Rivera’s kitchen, at least, robots won’t be replacing people — they’ll just help them do their jobs better. That might be an exoskeleton to help workers unload heavy boxes of produce, or Roombas vacuuming up at the end of the day.

Possibly the best leftovers plate we’ve ever seen.

For restaurants, food delivery is a blessing — and a curse. It can increase overall sales, capturing people too lazy to get off the couch, but on the flip side it also reduces foot traffic. Heimendinger made an interesting comparison: “The trend in ordering delivery from restaurants follows a little bit the trend of going to the movies,” he said. People are streaming movies at home instead of going to theaters, and are ordering food delivery instead of dining out. “This will change some of the function of a restaurant’s physical form,” predicted Heimendinger. He thinks it won’t be just a space to eat food: it will also, more than already, be a place to relax, meet up, have a meeting, etc. Which changes the business model. “It turns it into a community space,” he explained, “Not just a dinner table.”

Addo is certainly more than just a dinner table. Based on the interest we saw last night in addo’s far-ranging dining experience and community vibe, I wouldn’t be surprised to see a lot more incubators popping up in the future.

Want to hear more of Chef Rivera’s perspective on how technology can help lighten the load of restaurant workers and open up new revenue streams? He’ll be speaking at the Smart Kitchen Summit in Seattle on October 8-9th — get your tickets now! 

We ended the night by making some crepes with the Hestan Cue!

Spyce Restaurant Review: Delicious Bowl Food Made Me (Almost) Forget The Robots

Last month, I visited Boston with my sixteen-year-old son.

After a few days of site seeing in a city rich with American history, I asked him what his favorite thing he’d seen so far.

“The robot restaurant,” he said without hesitation.

So much for the founding fathers. Instead of Faneuil Hall or the site of the Boston Massacre, my son thought the coolest thing in Beantown was a restaurant where bowl food is assembled by robots.

I can’t really blame him. Overall it was a fun experience, and the food was tasty. Here’s a quick review of my son’s first (and favorite) robot restaurant.

Walk In and Order

We visited on a Sunday afternoon. Walking in, the first thing I noticed was not only were there a row of robots, but a couple of humans are running around helping out behind the counter.

Spyce’s human and robot workers

The place was crowded. People were ordering food at the kiosks, and most tables had people scooping food out of bowls.

There were a row of kiosks that are immediately visible when you walk in. A manager type was hanging out near the front wearing a suit, ready to help if we had any questions. We decided to try it by ourselves.

Ordering a meal at the Spyce robot restaurant in Boston.

The kiosk ordering experience was easy. It reminded me of ordering a meal at Eatsa, in part because I was ordering bowl food, but also because it was an intuitive ordering experience that moved very quickly.

There were a number of basic bowls to choose from. Each bowl cost the same: $7.50. I picked a Thai bowl and my son an Indian bowl. Once your  bowl is picked, you can choose from a number of add-ins and sides such as soft boiled eggs, cabbage slaw, and pumpkin seeds.  After sides,  you choose from one of three drinks: iced tea, hibiscus ginger and tart kiwi limeade.

Meal Preparation

Once a meal is ordered, the information is sent to the robots.

Meal Preparation At Spyce Restaurant

After portions of veggies, rice and protein are scooped into one of seven woks, a dispenser scoots around and squirts sauce into the meals.  From there, the woks start spinning and cooking. Above the woks, signs tell you what’s cooking. A sign above one wok read “Now Cooking Thai Bowl for Mike Wolf”.

The food is cooked quickly (the restaurant touts each meal takes about three minutes). I chatted briefly with the manager, who told me during the busy lunch hours the robots can prepare around 150 or so meals per hour.  He said that while lunch is their busiest meal, they serve around 40 or so meals per hour during dinner.

Final Prep and Serve

After a couple of minutes, the wok tips over and drops the steaming hot food into a bowl.

Adding Garnishes At Spyce Restaurant

From there, bowls are picked up by a human who adds garnishes and the sides. The experience wasn’t all that different from watching over a a Subway sandwich being made (minus the processed ham and turkey).  I chatted with the prep server, who politely answered all my questions.

The Meal

The meal is served in a paper bowl with a lid on it. On to is the customer’s name and bowl type printed on the lid paper. My bowl read “Mike Wolf” and it had the world “Thai” in small print.

The food was excellent. My meal, the Thai bowl, was yummy, but my son out-ordered me with the Indian bowl. Perfectly seasoned chicken, peas lathered in tikka masala sauce, potatoes and brown rice. Delicious.

The drinks weren’t as good as the food. Since the customer dispensed them at a drink station, I tried them all. I liked the limeade the best, but overall none were outstanding.

But this was ok because the food was tasty and cheap, and the order and preparation experience was quick and fun.

“I’d eat there every day,” my son told me as we left.

Paul Revere never stood a chance.

How An Obscure Academic Project May Have Just Started A Kitchen Robot Revolution

Imagine it’s 2031 and you’ve sat down for dinner with your family.

It’s middle of the week so tonight’s meal is nothing too ambitious, mac and cheese or fajitas. As is the usual routine, you catch up with the family and share a few laughs until the meal is finally served, at which point everyone loads their plates and starts chowing down on what turns out to be a tasty dinner (the third one this week!).  Soon your youngest -- the finicky one -- asks for seconds.

Congrats parent, another successful meal, but don’t spend too much time patting yourself on the back because here’s the thing: Neither you nor your significant other spent any time preparing tonight’s dinner.  Instead, tonight’s dinner -- and every dinner this week -- was prepared in its entirety by a robot, the very same robot who is now in the kitchen cleaning up after dinner and preparing dessert.

Futuristic? Yes. A science fiction movie cliche? Definitely. But the above scenario may also be a very realistic possibility in large part due to an obscure research project involving 32 GoPro adorned home cooks making dinner.

Creating A Technology Big Bang

With any technology that changes the world, there’s almost always a research breakthrough or two that helps unleash innovation. In today’s world of AI and robotics, most experts would agree that one of these technological “big bangs” was a 2012 ImageNet Challenge research team led by the University of Toronto’s Geoff Hinton.

ImageNet is a crowdsourced database of millions of annotated images. The accompanying ImageNet Challenge is an annual contest where teams of researchers in the area of machine vision come together to pit their machine vision algorithms on the ImageNet dataset and against one another to try and achieve the highest degree of accuracy.

Hinton’s 2012 team had what is widely believed to be a breakthrough in AI research by utilizing deep learning techniques to achieve much greater accuracy than before (85%).  Since this breakthrough effort six years ago, there’s been leaps forward each year -- today’s ImageNet Challenge teams routinely achieve 95% accuracy, better than most humans --  helping to drive significant progress in all corners of the AI world from autonomous driving to augmented reality to industrial and consumer robotics.

All of which brings us back to the kitchen.

And Now Into the Kitchen (The Epic Kitchen)

Now, a group of research academics is trying to create what is the equivalent of an ImageNet for the kitchen. Called EPIC-KITCHENS, the project is an ambitious effort to capture people performing natural tasks in their home kitchens like cooking, cleaning and doing laundry and then release the resulting millions of annotated images into the wild. The ultimate goal behind EPIC-KITCHENS is to create an open dataset about kitchen-centric objects, behavior, and interactions upon which researchers across the world can then focus their deep-learning algorithms on in the hope of advancing artificial intelligence in the kitchen.

Why the kitchen? According to the study’s lead, Dr. Dima Damen, the kitchen is one of the most complex environments in everyday life for artificial intelligence to master because it involves so many tasks and actions.

EPIC-KITCHENS 2018 TRAILER

“The most challenging type of object interactions tend to be in our kitchen,” said Damen in a phone interview I conducted last month. “We’re doing lots of tasks, on short notice, we’re multitasking. We might be adding something to our meal and moving something around. That makes the kitchen environment the most challenging environment for our types of perception.”

Damen, who is with the University of Bristol in the UK, partnered with researchers at the University of Toronto and Italy’s University of Catania to bring the project to life. The project took about a year to complete and involved a panel of 32 home cooks across ten nationalities in four cities in Europe (United Kingdom) and North America (Canada and US). To capture their activity, each participant mounted a GoPro on their head and went through 1-5 hours of preparing meals, cleaning and whatever else came naturally.

“We gave them a camera, sent them home, and said just record whatever you are doing in your kitchen for 3-5 days,” said Damen.

From there, the participants watched the video and narrated their videos so researchers had an audio track from which to manually annotate the atomized images -- 11.5 million in all- captured in the 55 hours of video.

The result is a massive database its creators hope will help researchers in training their AI systems to better understand the kitchen. Like ImageNet, the creators also hope to foster competition with challenges and will track the progress with online leaderboards.

The data itself is something many will find somewhat mundane:

Distribution of actions in kitchen. Source: Epic Kitchens

The above distribution of annotated actions and objects are what you would probably expect: a really long list of things -- like vegetables, kitchenware, spices -- found in the kitchen. Same for actions. The above distribution breaks down pretty much all the verbs we perform in the kitchen such as put, take, twist and so on.

And that’s the point, at least if you’re a researcher hoping to train an artificial intelligence system. Just as this type of granular data helped ImageNet Challenge teams achieve a 95% accuracy rate with their software, the EPIC KITCHENS team hopes to reach a similar level of accuracy. By helping these systems understand what everyday objects are and how people manipulate them in a series of actions every day to do the basic functions of like in our kitchen like cooking and cleaning, the EPIC-KITCHENS data and what evolves out of it can provide a foundation upon which technologists can eventually create robots that act like humans and perform human-like functions in the kitchen.

The result could be an explosion in innovation in spaces like augmented reality, personalized food identification apps and, yes, cooking robotics. And while a fully-functional Rosie the home cooking robot could be the ultimate end-result of this research a decade from now, chances are we’ll see much more evolutionary improvements between now and then in the form of smarter appliances, more capable virtual assistants and more immersive guided cooking experiences.

And oh yeah: if you’re the type who wants to keep the robots out of the kitchen altogether, don’t worry. One of the biggest challenges with machine understanding of food is that the three-dimensional human comprehension of taste, smell and texture is extremely hard to replicate with machines. Add in the difficulty of AI to understand context and it makes me think that while we may eventually get to cooking robots, they may only be average cooks at best.

The real artists, the chefs -- whether home based are on TV -- are probably safe from the robot invasion.

Probably.