Self-Learning Cooking Robot Swarm ChefSwarm Trials at Tokyo Michelin Restaurant: Mastering 120 Dishes in 3 Months
ChefSwarm system consists of 5 collaborative robots that autonomously mastered 120 dishes within 3 months through observing human chefs and repeated practice.
Japan's FANUC and Michelin three-star restaurant Ryugin's ChefSwarm project released interim results on February 27. ChefSwarm comprises five specialized cooking robots — a prep-cutting robot, a wok robot, a steaming robot, a seasoning robot, and a plating robot.
The system learns through an "observe-imitate-feedback" cycle: first recording human chefs' complete workflows through multi-angle cameras, then decomposing movements into executable robot motion instructions, and finally using tasting sensors (electronic tongue) and visual evaluation systems to score finished dishes and adjust parameters.
After 3 months of training, ChefSwarm has mastered 120 dishes, 30 of which meet Michelin star standards. Head chef Kazuhiro Honda commented: "The robots' heat control precision exceeds most human chefs, but there's still a gap in seasoning creativity."
Disclaimer
Content is AI-generated. Do not use it as a basis for real decisions. Do not cite it as factual reporting.