Chef Robotics raises $20.6 million to continue building AI robot arms

As the food service industry continues to struggle with labor shortages, a Silicon Valley startup has put together artificially-intelligent robots that are ready to clock in. 

Chef Robotics has raised $20.6 million in a Series A funding round led by Avataar Ventures, the company told CNBC. The round includes an additional $22.5 million in equipment financing debt to keep building out the startup’s robotic meal assembly systems.

CEO Rajat Bhageria and his team founded the company in 2019 with the goal of building robots that can help make up for a lack of human workers in food processing plants, which often see high turnover due to the nature of the job, forcing a reliance on temporary workers.

“There’s rooms of hundreds of hundreds of people, they’re just scooping food for eight hours a day in a 34-degree Fahrenheit room,” said Bhageria, whose startup is based in San Francisco. “It’s a lot more manual than even we had expected.”

The robotic chefs are long arms with multiple joints that are wrapped in a protective covering. The arms hang from metal racks that sit next to food-assembly lines. This allows the arms to scoop up different ingredients and spread them in prepackaged meal containers similar to how humans workers might. The robots are particularly suited for joining assembly lines where humans have to do repetitive motions.

The variable nature of food preparation has made it difficult for the industry to introduce more automation in the past. The difference in food textures makes it tough for machines to serve the correct portions of food, Bhageria said. Part of the problem stemmed from a lack of training data to teach machines how to handle different foods. A typical automated dispenser found in processing plants – like those that might fill bottles with sodas, for instance – don’t always meet the needs of Chef Robotics’s customers.

“I can’t just go to the internet and download training data – how do you manipulate a blueberry without squishing it?” Bhageria said. “That doesn’t exist.” 

To solve this, Chef Robotics is building its own training data to teach its robotic arms how to correctly serve food, Bhageria said. Already, these “chefs” have produced more than 40 million meals, Bhageria said. 

Automation in the food service industry isn’t a fully new concept, particularly at the restaurant level, where fast casual chains like Chipotle, Sweetgreen and Starbucks have been heavily investing in automation over the past few years to flip burgers, mix salads and more.

Chef Robotics’ focus on the pre-packaging assembly part of the industry has allowed the team to build what Bhageria called a “moat” of training data to add more ingredients as they expand their business to more clients. 

With the new cash infusion, Chef Robotics’ main goal is to expand to more plants and manufacturers in the U.S. and Canada. Bhageria said he also hopes to accelerate the robots’ imitation learning capabilities, meaning that the company could teach the robot arms how to perform an action by demonstrating it to the arm rather than writing code that programs it. 

WATCH: We see ‘extremely high’ appetite to adopt AI amongst manufacturers over the last 18 months: TCS