Like most chefs, Flippy is not afraid of hard graft. Since last summer that has meant 11am until 7pm shifts at Caliburger in Pasadena, California, as well as stints at Chick-N-Tots at Dodger Stadium in Los Angeles. These are long hours of repetitive work, where the boss demands absolute consistency. But you won’t hear Flippy complain.
Or say anything, in fact. For Flippy is a robot – a cloud-connected mechanical arm with 3D thermal scanners for eyes – that can flip burgers or fry 80 baskets of food an hour, monitor that food and even clean up afterwards.
Flippy was created by Miso Robotics and part-funded by the Cali Group, which is described by its CEO, John Miller, as: “A technology company that happens to sell cheeseburgers.” Cali creates new machines that it road-tests in its Caliburger restaurants across the world, into which Flippy will be deployed this year. And Flippy is not alone. Also in California, Bear Robotics has developed a self-guiding robot, Penny, which has so far served 40,000 diners.
Ten years ago, robot-chefs and waiters were pure sci-fi. Today, they are a reality, and at prices that make them a plausible investment. From May, Penny will be shipped on a subscription basis that offers “an immediate return”, says Bear Robotics’ chief operating officer, Juan Higueros. Miso Robotics says the cost to restaurants of leasing Flippy will be between $15,000 (£11,300) to $50,000 per year, depending on the complexity of the job it will be doing. Fully automated burger restaurants managed with minimal human oversight – where customers order at screens, pay electronically and eat food cooked and delivered by robots – are now a possibility.
The technology exists, it just needs knitting together cost-effectively and in a way diners buy into. In an industry keen to slash labour costs and increase profits, further automation seems inevitable. After an eight-month test at Kang Nam Tofu House in California, Bear Robotics credited Penny with driving a 28% increase in sales. Meanwhile, in US trials, the self-ordering screens that McDonald’s is currently installing at its 1,300 UK restaurants yielded a reported 30% rise in order values.
Personalisation is seen as key to the appeal of this technology (McDonald’s new screens allow you to customise your burger by, for instance, removing the gherkins), alongside its ability to reliably push so-called “upsells” (meal deals, extras, larger drinks) to customers. Soon when you log in to a restaurant app, it will be able to use your data and purchasing history to recommend dishes to you – factoring in everything from the weather to, if you are ordering at a screen that can read your face, your mood (as trialled by KFC in China).
“Right now people compromise on choice. We’re presented with a one-size-fits-all menu. Ultimately, technology can allow us curated choice. You love spicy food and chicken? Here’s the dish people like you mostly order,” says Tom Weaver, CEO at the hospitality tech company Flyt. He talks enthusiastically about the Helsinki pop-up Take-In, a full-service restaurant where diners ordered dishes in from various delivery services. “For the first time in a few hundred years, digital is allowing us to create new versions of restaurants.”
If high-street fast food is automation-ready, its impact on high-end restaurants is likely to be subtler and slower. “It’ll probably be a very long time before we see a Michelin star robot-chef,” says Robot Wars’ judge Noel Sharkey, emeritus professor of AI and robotics at the University of Sheffield. “Apart from the AI creativity gap, great cooking involves a subtle understanding of ingredients and delicate cooking that would be enormously challenging for robots. Placing fragile foods on plates would be incredibly slow.”
Currently, robots have limited functionality. A human needs to load Penny with plates that diners lift off themselves (it is more a runner than full-blown waiter). At the robot-powered US pizza-delivery company Zume, the fiddlier jobs, such as topping pizzas, are still done by humans. But Moley, which will launch later in 2019, offers an idea of the sophistication to come. Modelled on the movements of the 2011 Masterchef winner, Tim Anderson, its robotic arms are fixed over a stove and programmed to prepare dishes from raw ingredients. It is designed for domestic use, but a commercial version is planned that, like Suzumo’s sushi robots or the Foodini 3D food-printer, opens up the possibility of restaurant kitchens automating even highly technically challenging tasks.
For now, the kitchen technology that QSR Automations supplies to Mitchells & Butlers’ pubs or KFC is more prosaic. Rather than having a chef shouting out diners’ orders, the QSR system automatically distributes those orders across screens (multilingual for international brigades, with cooking pointers if necessary). The kitchen screens then tell chefs when to start cooking each dish or its components, so that your plate of fish, chips and mushy peas, or table 11’s order, comes together at the right time.
The restaurant industry is currently suffering a severe shortage of skilled chefs – a deficit of 11,000 to 2022, according to the skills agency People 1st. “Automation systems designed to make the process more idiot-proof are trying to bridge that gap,” says Stefan Chomka, the editor of Restaurant magazine.
In terms of simplifying or enhancing the customer experience, online restaurant booking has been the runaway success of the internet age. Automated at-table ordering and payment is seen as the next watershed, as pioneered by Wagamama and Wetherspoons. But, says Chomka, across the wider industry such digital technology is a “clunky” confusion of endless restaurant apps, QR codes and, if you want to pay without physically presenting your card, competing methods (eg ApplePay, Qkr!, PayPal). The restaurant industry is still awaiting its Uber moment: one universal order-and-pay interface to which diners can sign up once and use everywhere.
Until then, early adopters, such as the Manchester bar-owner Andy Smith, who uses the Ordoo app at his venue 33 Oldham Street, may well experience growing but “lower than expected” take-up. Undaunted, Smith remains an advocate for at-table ordering. “You don’t go to the pub to queue at the bar or peel away from good conversations to order. I once spent 90 minutes in a bar queue on New Year’s Eve and missed the countdown. It’s been a mission ever since to make that process easier.”
Splashy failures in restaurant tech tend to hog the headlines. The TGI Fridays drone that injured a Brooklyn Daily photographer; US burger chain Shake Shack having to retreat from plans for entirely cashless, kiosk-ordering sites. But, says Chomka, such technology could miss the mark if it only offers “solutions to problems that don’t really exist”. Recalling a brief fad for putting wine lists on iPads, he says: “People don’t want holographic representations of food telling them where the asparagus is from.”
“Payments being automatic in the ether” could be useful, he adds. But, in an expensive restaurant, “how quick do you need dinner to be?”
That attachment to the human art of hospitality will be heartening for staff who see automation as a threat to jobs. “Unite is concerned that without safeguards, any benefits will bypass workers in low-paying jobs, such as waiters,” says the union’s London regional officer, Dave Turnbull. “We’re urging MPs to ensure automation does not just benefit a wealthy few like previous industrial revolutions.”
The industry mantra is that automation will allow it to redeploy staff more effectively. Jobs will change, if not immediately decrease in number. “Self-order screens mean more demand for people front-of-house. Last year, we recruited 1,000 new managers,” says McDonald’s COO, Jason Clark. Caliburger has a programme to retrain staff as robot engineers or “chef techs”.
But ultimately the chance to reduce overheads by employing fewer people or at lower rates will be what attracts big brands to technology. Invariably, over the next 10 years, jobs will go in bricks and mortar restaurants, not least because automation will boost their rapidly growing rival: delivery.
Companies such as Domino and Just Eat have both experimented with robot and drone delivery, which is already legal in Reykjavík, Iceland. “Robot food deliveries are the future,” says Sharkey. “Small, 4mph ground-robots are safe and will work 24 hours a day.”
Combined with machine learning of anticipated order volumes and the minute-by-minute cooking data generated by systems such as QSR, that robot fleet could deliver almost instantly. “Restaurants [will] prepare food ahead of orders coming in,” predicts Just Eat’s director of engineering, Daniel Richardson. “I’ll be able to say: ‘Let’s have pizza’ to my family, my intelligent assistant will hear and 10 minutes later the food will be at my door.”
No one is putting an exact date on it yet. But in the dining room and on your doorstep, the robots are coming.
• This article was amended on 14 March 2019 to clarify the costs of leasing Flippy, after details were provided by Miso Robotics after publication. Also, an earlier version incorrectly said Miso Robotics was part of the Cali Group.
