A Bear Robotics robot delivers food to tables at Denny’s in Burbank, California. Technology promises cost savings, but it turns out the dining experience is all about feelings. — Photos: The New York Times
On a recent Thursday afternoon at a Denny’s in Long Beach, California, a four-foot-tall robot glided toward Diane Deconnick’s table. Its three tiered trays could have hauled a feast, but they were empty except for a bowl of soup and a fried chicken salad.
As the machine rolled to a stop, a scrolling digital display announced, “FOOD IS HERE.”
“It talks!” Deconnick said.
A Denny’s regular, she had been curious about the robot since it appeared a few weeks earlier. This was her first time being served by it.
Or partly served by it. A server had followed the robot. He took the food off the tray and put it on the table.
“What’s your name, robot?” Deconnick asked. “Lily,” the server answered. He’d named it himself.
Deconnick, who planned on tipping her usual 20%, said, “I like Lily. She’s a good worker.”
Robotic future?
For a decade, the promised era of robo-flipped burgers and automated baristas has always been just around the corner. But automation in restaurants, at least in the United States, remains a novelty. That is not because the robots and artificial intelligence assistants can’t hack it. By and large, the technology is there and in some cases has been for years.
As labour costs rise and robots grow more sophisticated, the cold economic case for automation may become irresistible. But in the meantime, our soft, irrational human emotions – how we react to these robot assistants – will play a decisive role in how much and how quickly they start to run our restaurants.
Even in a world of ghost kitchens, QR codes and contactless delivery, restaurants are still in many ways about feelings.
Michael Giebelhausen, a professor at the Wilbur O. and Ann Powers College of Business at Clemson University who studies the intersection of technology and hospitality, explained, “We should be thinking about not what jobs robots will take, but what jobs consumers will allow robots to do.”
Robots can be particularly useful for repetitive, manual tasks like carrying heavy tubs of dishes from the dining room to the kitchen.
So far, and perhaps surprisingly, much of the automation currently being rolled out in the US is on the human-facing, service side – robotic bussers, AI-powered drive-thrus and the ever-proliferating touchscreens.
But people come to restaurants to feel connected to other humans. They want to encounter people, not a chatbot, kiosk or mechanical arm. So successfully integrating robots is more than just an engineering challenge.
Giebelhausen has found, for example, that consumers prefer human chefs to robot ones, in part because they believe that humans cook with love. In a paper currently under review, he and his co-authors found that if consumers had a friendly text chat with the robot, that preference faded.
“The crux is, if you feel the robot loves you,” he said, “you allow the robot to cook with love.”
Human emotions are decisive for restaurant employees as well, though in their case, it’s less about feeling the love than assuaging the fears. An observational study conducted in Central Florida found that if restaurant management had consulted workers before bringing in a robot, they used it more effectively. At restaurants where a robot simply appeared one day with no employee input, workers were more likely to feel frustrated.
Mindy Shoss, an organisational psychology professor at the University of Central Florida who worked on the study, said that automation can make workers anxious about losing their jobs or even their entire occupation. “Given popular discussions about tech replacing workers, it is a common concern, especially when the organisation isn’t clear on why the tech is brought in.”
One reason they can be brought in is to do the heavy lifting.
Juan Higueros is the chief operating officer and co-founder of Bear Robotics, an American robotics company that participated in the UCF study and whose robot was gliding around the Long Beach Denny’s. He said the co-founders of Bear used their previous experience as restaurant owners to design service robots. Servers were walking and carrying way too much, and some trays on wheels could help.
Higueros said Bear’s data shows robots carrying thousands of pounds of food on behalf of human servers and runners. “Operators say, wow, my staff doesn’t have to do that anymore,” he said. “Robots play really well when doing mundane things people don’t want to do anyway.”
Of course, the replacement fears are not baseless. Peter Kim, chief technology officer of Navia, a service robot company, said he has seen an uptick in interest in California since the US$20 (RM82) minimum wage for major chain restaurants went into effect in April.
With robots, he said, “You don’t have to worry about sick days. They’re available 24/7 and 365; you don’t have to worry about weather or traffic – they just work.”
As labour costs rise and robots grow more sophisticated, the cold economic case for automation may become irresistible. But in the meantime, our soft, irrational human emotions – how we react to these robot assistants – will play a decisive role in how much and how quickly they start to run our restaurants.
Bear Robotics also chose to focus on dining-room robots because they offered a clear solution. The kitchen was much more complicated.
“In our own restaurant, we had four different languages being spoken. People are crowded trying to get everything done, rushing to get orders out and food on time to customers – back-of-the-house is a battleground,” Higueros said.
Gennadiy Goldenshteyn, the founder and managing partner of Dinemic Ventures and previously the head of global engineering at Yum Brands, a conglomerate that includes Taco Bell, KFC and Pizza Hut, said developing hardware for kitchen automation takes years, and faces logistical challenges like worker safety and sceptical franchisees.
As a result, even as chain restaurants look outwardly more futuristic, their kitchens work largely as they have for decades.
“It’s kind of like we made a self-driving car out of a Model T,” he said.
Help or hindrance?
Drive-thru ordering using AI was deployed by several large chains last year, though it showed mixed results and has been largely shelved for the moment. But recently, at a Rally’s in Orange County, AI was still in place. The program understood an order for chicken fingers clearly, though it didn’t respond to follow-up questions.
“It’s a little bit of help, more help than before,” said Alondra Villegas, the shift manager. But Villegas said customers were split about “half and half” on whether they liked the assistant. “She’s funny – sometimes she doesn’t answer customers, and customers get mad at her.”
Goldenshteyn believes that all of the order-taking automation, including kiosks and AI, will eventually be superseded by a familiar bit of hardware that consumers already have a strong emotional attachment to: the smartphone. Most major chains already have apps that allow customers to place phone orders with fairly little friction.
And while human chefs excel at creating new dishes, he said automated cooking is the future, whether or not consumers believe that robots can cook with love.
“Pepperoni doesn’t care if it was put there by a human hand or by machine,” he said.
Whether in the drive-thru line, the dining room or the kitchen, the key may be to stop designing robots that merely replicate how humans work.
Darian Ahler, a food tech consultant, said that a common mistake is to create a humanoid robot that performs a task that’s easy for a human to do – an arm making a latte, for instance.
If consumers prefer human cooks, and cooks worry about losing their human jobs, making a too-human robot may backfire. Instead, Ahler believes that automation “needs to act like a toaster”.
He helped develop the Autocado, an avocado-processing machine that is now being adopted by Chipotle. A waist-high metal box with a chute for cooks to dump in avocados, it does the tedious work of peeling, halving and pitting. “People don’t see this as taking jobs,” Ahler said. “They see it as assisting.”
Even though most restaurant kitchens work much as they have for decades, the fully automated restaurant, especially one focused on making a single type of dish, is closer to science fact than science fiction.
Benson Tsai, CEO and co-creator of Stellar Pizza, a fully automated pizza truck, said he was inspired by frozen pizza factories. “Walking into the grocery store, everything there is made by robots,” he said. “You can watch YouTube videos of pizzas being made by the thousands on conveyor belts. It’s a beautiful thing.”
A veteran of the electric vehicle industry and SpaceX, Tsai said his interest in automating restaurants began when he saw food costs going up. “If people could pay US$2 to US$5 (RM8 to RM20) less for their food, that’s much more beneficial than the few jobs we might be removing,” he said.
Operating in the Los Angeles area, the Stellar truck turned a profit selling 12in pizzas for about US$10 (RM43), Tsai said. In March, the company was acquired by Korean conglomerate Hanwha Food Tech. The truck is no longer in use, but Hanwha plans to open full-service restaurants in Los Angeles.
Part of the magic, Tsai said, was hiding the robots from customers. “I’m a firm believer that automation should be felt and not seen. I want them to feel it in their wallet and taste the higher quality of ingredients.”
The truck did employ one human, however: someone to hand the pizza to the customer. Tsai pointed out: “Humans still do one thing very well, and that’s smiling.” – The New York Times
