Two years later, a Flippy Fry Station—equipped with fry baskets instead of just a spatula—debuted at Dodger Stadium and the Arizona Diamondbacks’ Chase Field to deep-fry 31,000 pounds of chicken tenders and Tater Tots for baseball fans. An updated third-generation robot is now able to fry and portion more than 40 menu items and reduce staff interactions with the machinery by 90%. According to company filings, Miso sees a potential $4 billion revenue opportunity with Flippy’s automation tools.

“He just does it twice as fast, and he does it perfect every time,” Miso Robotics CEO Rich Hull told Fortune. “And he never comes in sick and doesn’t take a smoke break.”

The emergence of Miso Robotics comes as the global restaurant automation market is expected to grow to $28 billion this year. Robots are now able to make fried rice and pasta in a wok, prepare salads, and cut and core avocados used to make Chipotle’s guacamole.

Miso has continued to expand its own operations. It finalized the acquisition of Zignyl, an AI-powered restaurant operations system, at the beginning of the year, the company told Fortune exclusively. The integration of Zignyl will allow restaurant operators to essentially run Flippy, as well as point-of-sales systems, labor scheduling, and payroll, through an app. Miso did not disclose the details of the deal.

For all Miso’s efforts, the startup has hit snags. Miso generated about $385,000 in net revenue in 2024, down from about $493,000 generated in 2023. It ended partnerships with CaliBurger and Panera. As of the end of 2025, Miso has 14 Flippy units in White Castle and Insert Coin. Two years ago, the company reported 17 units across White Castle, Jack in the Box, and other partners. Reddit users who say they are investors in Miso have sounded the alarm on the company’s cash-flow negative status and failure to grow its number of active units.

Hull said 2024’s flat revenue was the result of Miso developing and launching the third generation of Flippy meant specifically to create scale.

“The first two generations of all disruptive technology products are necessary experiments, but it’s generally the third generation of anything that scales, and Flippy is no exception,” he said.

But the challenges Miso faces are reflective of an automation push that has yet to prove its worth in the eyes of some economists. Intended to address restaurant labor shortages and improve efficiency, machinery automating some tasks have shown themselves to be expensive to build and maintain, let alone scale widely across the food service industry.

Kernel, a vegan fast-casual restaurant in Manhattan known for its Kuka robot arm that prepped dishes, closed its doors within a year of opening and rebranded as Counter Service, a more human-powered, quick-service sandwich joint. In November, Sweetgreen sold Spyce, the division that developed the chain’s Infinite Kitchen automation technology, to Wonder for $186 million. Sweetgreen said at the time the sale was part of a strategy to refocus on profitability.

As AI comes under increased scrutiny, with conflicting narratives about its impact on productivity and employment, some experts say physical automation is going to be an even harder sell. 

“It’s not about, necessarily, technical possibility,” Ioana Marinescu, an economist and associate professor of social policy at the University of Pennsylvania, told Fortune. “It’s about economic value, whether this particular technology is worth it in the current economic context.”

The rise of restaurant robotics

The pandemic exacerbated a labor shortage crisis in the restaurant industry that laid the groundwork for back-of-house automation to gain popularity. According to the National Restaurant Association, full-service restaurants lost nearly 3.7 million jobs in the first two months of COVID, and it has yet to fully recover, with employment remaining 210,000, or 3.7%, below pre-pandemic levels. Fast-casual and quick service restaurants have recovered and sit at about 79,000 jobs, or 2%, above pre-pandemic levels.

“We can’t hire enough. Nobody wants to work,” Hull said. “And everybody originally kind of thought that was temporary, but in reality, it just accelerated something that was already happening.”

The Trump administration’s immigration crackdown has also had a disproportionate impact on restaurants and other hospitality sectors as undocumented workers skip work out of fear of arrest or increased scrutiny by authorities.

According to Hull, short-order cooks are among the hardest to retain in fast food, given how messy and potentially dangerous fry stations can be. Turnover can be expensive for restaurant owners, costing them more than $2,700 per hourly worker, according to data from Black Box Intelligence. About 35% of that sum comes from providing new worker training. That’s on top of rising labor costs, which 98% of restaurant operators identified as an issue for their business, according to the National Restaurant Association’s 2024 State of the Restaurant Industry report. 

Enter restaurant robots, machines promising to alleviate unwanted or menial tasks for kitchen staff while increasing order efficiency and creating consistent quality. Analysts like Bank of America’s Sara Senatore said restaurant robots won’t significantly displace their human counterparts, but rather improve their well-being and increase the likelihood of their staying in the job longer.

“You make the human existence way better,” Hull said. “You drive more revenue for the restaurant operator, and you allow people to stay longer in these jobs, have careers, get promoted, get paid more.”

Hull said the demand for robots like Flippy has been “massive” as restaurants grow more comfortable with the idea of back-of-house automation. White Castle has previously said it wants Flippy installed in one-third of its 350 locations.

Hull noted the cost for these restaurants to install Flippy is low, about $5,000 per month, or “less than you would pay the equivalent human.” He added the robot can double the output of what a short-order cook would be able to serve.

The slim chances of a robot revolution

Fulfilling this high demand is where economists warn things go sideways. University of Pennsylvania economist Marinescu said there’s a reason we haven’t seen widespread use of robotics in service industries.

“We’ve been at this task for centuries, of trying to automate the physical stuff that people are doing,” she said. “So it’s not for lack of trying.”

A January 2024 study from MIT CSAIL, MIT Sloan, the Productivity Institute, and IBM’s Institute for Business Value found that more than 75% of the time, it is cheaper for companies to continue to use humans than to automate jobs with AI. The study authors give an example of a small bakery, calculating that if that bakery had five workers getting paid $48,000 annual salaries, a visual inspection task assigned to AI that reduced a baker’s work by 6% would save the bakery $14,000 per year. This sum is significantly less than the cost needed to develop and maintain this AI system, the study said.

A large part of these expenses come from AI-powered hardware, which is highly specialized and harder to scale. Marinescu and colleague Konrad Kording found in a working paper published by the Brookings Institution in November 2025 that costs to develop and build robots are declining at a slower rate relative to the costs of AI software. ARK Investment Management projected in 2019 industrial robot costs would decrease 50% to 60% by 2025. AI expense management platform Ramp noted in April 2025 companies are now paying $2.50 per 1 million tokens compared with $10 one year ago, a 75% drop in costs.

“While surely possibly you can squeeze a tiny bit more, maybe, out of this new technology,” Marinescu said, “it doesn’t seem very plausible that there will be a huge win in productivity.”

Hull, for his part, said the cost to produce a Flippy arm was $80,000 five years ago but today costs 80% less.

Ajay Agrawal, professor of entrepreneurship at the University of Toronto’s Rotman School of Management, told Fortune, not only are restaurant robots expensive, they also require full buy-in from restaurants to be fully effective.

“To really get the benefit of robots or artificial intelligence, you need to redesign the whole system, rather than just including one robot to do a particular thing,” Agrawal said.

Industrial kitchens, for example, are small, tight spaces that can easily become less efficient if a human has to navigate around a large machine. Operations become most successful when they are vertically integrated, Agrawal suggested. Unless a company can create a unified system, the success of automation will be significantly limited.

“The real winners are going to be the ones who recognize, ‘Okay, these robots are very accurate, they’re fast, and we should redesign the restaurant around the capability of the robot, and then have people do other stuff,’” he said.