Miso Robotics started as an unproven idea. John Miller, chairman and chief executive officer of the parent company of fast-food chain CaliBurger, and venture capitalist Buck Jordan wanted to bring innovation into the kitchen. They looked to Sinnet, Zito, and Anderson to engineer a product.

Thrilled at the prospect of devising a disruptive technology, Sinnet discovered that the artificial intelligence technologies he learned in Ames’ lab—computer vision, machine learning, and forecasting—could be adapted for restaurant robots. He needed patience, a willingness to understand the industry, and some Caltech ingenuity. “We brought in Techers as employees and interns when we were starting from zero,” Sinnet says. “They brought so much creativity early on. I don’t think we could have gotten where we are today without all their help.”

The first iteration of Flippy debuted at CaliBurger in 2018, less than a mile from the Caltech campus. Part worker and part spectacle, Flippy cooked patties from a demo kitchen designed to attract customer interest. The robot used computer vision to identify food, a thermal camera to track the temperature of the patties, and control theory to pick up a spatula, flip the burgers, and clean the grill. The newer version of Flippy, which is configured for frying food, builds on these technologies to cook several types of food at once and to forecast customer demand.

Sinnet spent considerable time at CaliBurger, searching for ways to improve Flippy’s operations and functionality. In the process, he befriended the CaliBurger staff and saw them grow fond of Flippy. They gave it their own nickname and one joked with Sinnet, “Don’t take away my robot.” No one expressed concern about robots taking their jobs, Sinnet says.

He attributes this harmony to two factors. First, robots cannot compete with human workers—at least not yet. “A human can learn and perform hundreds of tasks in a day,” Sinnet says. “It takes comparatively longer to teach a robot to do just one.”

Second, restaurant workers have enough to do. As employers searched to fill open positions, Miso Robotics test sites increased kitchen productivity by 30%, according to the startup. Staff has more time to engage in tasks that are more meaningful to humans, such as interacting with customers, Sinnet says.

“My work has always been about using technology to advance human welfare. Let’s solve the easy and mundane problems and then focus on the harder ones that we are passionate about.”