Can a robot rule the restaurant? Cali Group says yes
Cali Group CEO John Miller and one of his newest babies, the updated, rail-based Flippy. The new model can move between multiple stations to flip burgers and fry fries.
Photos by Scott Witter
The morning I was to meet Flippy, the burger-flipping and fry-frying robot, I landed smack dab in the middle of a chaotic breakfast rush at my hotel.
It was rough. A guy near my table asked three times about his oatmeal in the span of 45 minutes—yes, oatmeal. The counter staff was exceptionally slow and got my order wrong twice, and then gave my tabletop number to the wrong person. And when I wanted to add toast to my omelet—a common addition I’d assume—we all had to wait until her manager could find her the toast button.
Then, there was that rock star kitchen manager, glistening with sweat, running around with food, picking things up on his way back and fielding questions from confused or annoyed patrons. Amid all this, the custom omelet bar employee was just standing there looking exceptionally bored.
What if there was a better way? What if I could walk up to a kiosk that recognized my face, recommended a good omelet and already knew I loved toast? What if that kiosk told a robot in back to make my omelet? And what if an employee just had to bring out the food and chat about the weather? What if instead of sweating through his shirt, the kitchen manager could focus on making sure the robot was stocked and every plate went out looking great?
That’s exactly what John Miller, the founder and CEO of CaliBurger and parent Cali Group, is working on. The restaurant industry is on the verge of a major change.
“The time is now to invest in innovation and the reason why is there’s two intersecting trends. One is the restaurant industry has come under attack from increasing costs,” said Miller. “We see that it’s wide open for someone to own the future, what I’ll call the restaurant operating system.”
Miller was once a legal and finance guy at a public pharmaceutical company, and he literally wrote a book on the nanotechnology business. But in a tale as old as time, he got the restaurant bug, tapped his bank account and started an American burger concept in the middle of China.
“We were doing a lot of business in Asia. I thought it’d be fun to open up a restaurant business in Asia with some friends; we realized quickly how hard that was,” said Miller, describing the first CaliBurger location. “But we also realized how much opportunity there was to deploy tech into the restaurant industry. Basically, I decided to dedicate the last eight years to apply technology to the restaurant business.”
Now at more than 32 locations around the world, CaliBurger serves California-style burgers, fries and chicken sandwiches. It also serves as a lab for all the technology under Pasadena, California-based Cali Group, also the parent of several restaurant technology companies.
Among the forward-looking companies in the mix is Kitchen United, the ghost kitchen developer founded to lean into the growth of delivery with multiple kitchens. It aims to help restaurateurs and brands grow off-premises operations without front-of-house staff or traditional real estate. Then there’s PopID, the facial recognition software that recognizes customers for loyalty or payment at a kiosk or counter. And Cali Group has already spun one company out to the public markets: the esports-focused community platform Super League Gaming.
But the most viral and most futuristic company in the stable is Miso Robotics, the creator of Flippy and Miso AI, the software that keeps the robotic arm flipping burgers and dropping baskets of fries, tots, chicken wings and other staples into the fryer.
The machine learning software and the robotic arm work in tandem to monitor the burgers and knows exactly when to flip, when to add the cheese and when the nearby worker needs to get a bun or more burgers ready to go. Recently, Flippy took over the fry station as well, timing exactly when to remove the basket of fries, chicken tenders or tater tots. It’s currently pumping out basket after basket of chicken and fries at Dodger Stadium in Los Angeles with a total of five operating outside the Miso Robotics lab.
The chief goal: get people away from the dirty, smelly jobs in foodservice and by doing so, drive down costs associated with labor, especially turnover.
“Our big focus now is how much we can reduce on the income statement by reducing turnover. The cost of running a restaurant right now is driven much more by turnover than it is by rising wages,” said Miller. “We’re seeing the robotic systems improve the operating cost in two ways. One, people stay longer because they enjoy their job more when they don’t have to do the hard jobs of working on the grill or the fryer. And they can spend more time doing what people are good at.”
The cost of turnover in restaurants, according to research firm TDn2K, is $1,800 for hourly workers, $10,300 for managers and nearly $14,000 for general managers.
As any restaurant owner knows, turnover is extremely problematic.
The Bureau of Labor Statistics reports that between 2015 and 2017 hospitality and foodservice turnover reached 70 percent. A new survey by the bureau is coming in 2020, but restaurants are seeing levels above 100 percent routinely and as high as 150 or 200 percent in especially difficult labor markets.
Rob Anderson, co-founder and head of mechanical engineering at Miso Robotics, said one of their clients has 8,000 foodservice job openings.
“I think about why is that an issue. What I want our technology to do is make the restaurant a better place to work, where employees don’t have to do the menial, dangerous tasks, but be engaged with the customer engagement,” said Anderson. “And open up the opportunity for career growth in restaurants.”
Instead of slaving away at the hot grill or counting oil burns at the fryer, those “chef techs” set up the robot, provision it with the patties or fries throughout the day and make sure everything is going well. And currently, they’re watching for issues with the nascent system to make its performance flawless. The robot doesn’t get sick, doesn’t take a job down the street for 50 cents more, and when one robot gets better, they all do.
“Let’s say you’re a 500-location chain, and you find a better way to make nuggets, we can upload that to all your locations,” said Buck Jordan, co-founder and CEO of Miso Robotics.
Beats writing a new training manual and shipping it out, eh?
From left, Miso Robotics leaders Rob Anderson, Buck Jordan and Ryan Sinnet talk through the logistics of the mobile Flippy cart, designed to move for cleaning and easier kitchen operations.
The messy work of innovation
When you watch a video of Flippy, or hear the technologists talk about how proficient the robot is at making some 2,000 burgers a day, or 80 baskets of fried food an hour, it’s easy to forget how brutally hard this kind of innovation really is.
“It’s really cutting edge work, a lot of computer vision, data science, hardware and electrical engineering,” said Jordan. “It’s a whole concert of people required as opposed to a traditional software business where you only need two kinds of people. Here, you need a whole bunch.”
When you walk into the Miso Robotics office, the first thing you’ll see is the electric scooter stand (so Pasadena), and then the rows of computers out in front of a pile of computer parts and a few robotic arms. It’s chaotic, supremely nerdy and totally awesome.
The bright minds behind those computers, sharing code or talking through some issue with the computer vision, are generally young and laser focused. While Franchise Times was there taking photos, they worked through many disruptions, maneuvering however they could to get back to the job at hand. Many came from the California Institute of Technology, conveniently located just up the street from the Miso headquarters and an ideal feeder for the companies within Cali Group.
Co-founder and CTO Ryan Sinnet said the chief commonality among those bright minds is the chutzpah to build the future of restaurants.
“We need people with really high aptitude. These problems are really hard and sometimes unsolved, so we need people who are bold enough to even want to try,” said Sinnet.
Whip-smart innovators are one thing, but having a restaurant group willing to play guinea pig is a perfect partnership.
“This is really, really difficult, groundbreaking work. It’s the most cutting edge advanced robotics I’ve ever seen. So it takes a dedicated group of talented engineers and some forward-thinking QSR restaurant partners,” said Jordan. “What we require is someone to work with us and understand the value we bring.”
He said having Miller’s overall vision of the restaurant operating system linking CaliBurger, Kitchen United and Miso Robotics in the same building is maybe the ideal crucible for this kind of work—and PopID is just a few blocks away.
“John had created this Willy Wonka’s Chocolate Factory of restaurant innovation. In a small radius, you have several groundbreaking restaurant concepts. So what’s beautiful, you have a lot of synergistic companies,” said Jordan.
It also keeps the work focused on the real problem at hand. As Anderson put it, there’s always “this dilemma of looking for a solution without understanding the problem.” A group of engineers with the best intentions can spend far too much time working in the wrong direction.
“In a special way, we built Miso from the franchisee perspective. We’re not some tech guys coming in to tell you how to run the restaurant, we have CaliBurger and Cali Group as our founding partners,” said Jordan. “So we had them making sure we were building the right product at the right time.”
The process of getting the product right, however, can be a bit deflating at times.
“We had our first burger flipped in two months. We all went in this tiny garage, the machine flips a burger and we all kind of high five and cheer,” said Jordan. “Then John said, ‘That’s great, but that’s not food safe.’”
That one observation led the team to develop a food-safe sleeve to wrap around Flippy’s robotic arm.
“If we didn’t have that kind of input, we probably would have spent nine months engineering the wrong thing,” said Jordan. “Because of that constant feedback, we can do the right thing.”
Flippy is working right now in a few locations, and the first “CaliBurger 2.0” just opened in Florida, remodeled to be the ideal home for Flippy version three. Then, the company will bring Flippy to the Seattle market, plus a newly renovated home restaurant in Pasadena and to Washington, D.C., all in 2020. But as the futuristic dreams collide with reality outside the Miso test kitchen, there have been some pivots.
The first Flippy in operation was quickly shut down after a few days. Not because it was too slow or too fast, but because of an archaic zoning issue. Part of the kitchen space the robot inhabited was zoned for salad. And when flagged, the city council wanted to learn more about just how safe the robotic arm capable of bending steel beams really was. After a few days, the council was satisfied and Flippy was back at work.
Another learning involved an early version that was bolted to the floor. If it malfunctioned, it wasn’t only annoying but the annoyance was immovable. So the company pivoted to the cart-based arm as seen in Dodger Stadium. And at the Florida location, the company and franchisee renovated the restaurant so there’s nothing out in front of the cooking equipment. Instead, Flippy operates on a rail system against the wall. Cooking backwards, it seems, is not an issue for Flippy. Sinnet said the goal is to get Flippy into as many restaurants as possible, without the need for much, if any remodeling.
“We’ve put all our time into building these flexible products. Now, one arm can do multiple stations. Our future is having it on a rail to work on one, two, three or more stations,” said Sinnet. “And making them not require a computer science degree to operate.”
That’s another must-have for the robotic QSR future because even if turnover sinks dramatically, it will never go away.
“We’ve really developed on the hardware side a solution that is easy to deploy and our aim is to work with existing restaurants, without the big barrier of building restaurants,” said Anderson. “We’ve designed the system to be quick to deploy and set up in a day, and train people in like 10 minutes to understand it.”
That, he said, is what sets Flippy and Miso apart from other automated foodservice models.
Competition and cost
There is a diverse range of models out there, such as a burrito-making vending machine designed with gas stations in mind. There’s Zume and Ono Blends, two concepts making meals out of automated food trucks—pizza and healthy smoothies, respectively. Then there’s the big Rube Goldberg-esque grain bowl-making machine seen in Boston-based Spyce, thought up by MIT students and with a menu by legendary chef Daniel Boulud.
Another burger bot is at the core of the restaurant Creator, which makes burgers to order along a conveyor belt, even grinding the meat to order. These are all boutique solutions, requiring what amounts to a computer scientist, an electrical engineer and a technician on staff to keep the machine going.
Mechanical arms, on the other hand, are ubiquitous. They’re designed to work for eight years in continuous operations without maintenance, and they’re easier to understand. There are more simple versions, like the automated barista bot of Café X that picks up and moves your coffee around a workstation as it works. The difference with Flippy is in its visual acuity and machine learning—it gets better where a simple program cannot.
And none of those systems, aside from the pre-programmed bots, are exactly plug-and-play, requiring intense setup and calibration before operating and needing arduous upkeep.
But what about the ultimate question: How much does it cost to put a burger bot in your kitchen? Well, it’s coming down.
At first, the arms were about $60,000, with the software, the cameras and everything else on top of that. That’s a lot more than even a fully automated oven, but prices are dropping.
Arms capable of taking on Flippy’s duties have dropped to about $20,000. Sinnet said they’re looking to halve that cost.
“We’re figuring out how to get the bill of materials under $10,000 so it can basically be given away, that’s what we really need to nail,” said Sinnet. “In under two years we’ll be at that lower cost.”
At that point, he said, the company will be able to set up the bot and charge monthly just like any other software as a service. And by then, noted Miller, Flippy will be ready to plug into all of the disruptive restaurant trends happening out there.
“You’ll get a lot more attention from people when robots are cooking the food,” said Miller. “We also believe that the world is changing as fewer people come into restaurants, so the experience has to be engaging and there will be a lot more delivery. With some form of autonomous vehicle to bring the food home, be it drones or driverless cars.”
Flippy at the fry station inside a Dodger Stadium concession stand.
Flippy Goes Public
Flippy and Miso Robotics are heading toward the public markets under the still-novel Regulation A+ path that makes investing in the company within reach of everyday investors. Right now, the company is taking commitments from investors as it works through the Securities & Exchange Commission documents and procedures.
Co-founder and CEO of Miso Robotics Buck Jordan said the company is in a special position because of the viral nature of the burger-flipping robot.
“We have 10 billion unique media views, that’s an insane number, we could start a shoe brand,” said Jordan.
Don’t hold your breath for Flippy Flops, but he said the new investment tactic is ideal for a loved brand such as Miso. And as a former venture fund finance guy, Jordan said it’s a way to let the company thrive in a new era of investing.
“Venture is becoming democratized. This is a very stodgy industry that only allows the 1 percent to invest, so why shouldn’t everyone have that opportunity?” said Jordan.