Xianglu Robotics Launches Three AI-Powered Cooking Robots
Xianglu Robotics has launched three new products at the 2026 World Robot Conference: the CookingMuse multimodal AI base model, the 3K Vision AI stir-fry robot, and the "Xianchao Fangzhou" mobile unmanned kitchen solution. The 3K robot uses industrial cameras and an AI model trained on millions of dishes to dynamically adjust cooking parameters in real-time based on visual feedback. This development showcases the integration of computer vision and embodied AI in commercial kitchens, shifting automation from rigid pre-programmed recipes to dynamic, real-time adjustments. It could significantly improve operational efficiency and consistency in the food service industry by handling variations in ingredients and preparation. The 3K robot features three 40-megapixel global shutter industrial cameras capturing video at 30 FPS, allowing millisecond-level adjustments to heating power and cooking time. Meanwhile, the "Xianchao Fangzhou" acts as a mobile, self-cleaning unmanned restaurant that manages the entire workflow from ingredient storage to serving without requiring exhaust construction.
## BACKGROUND
Traditional cooking robots rely on pre-set programs, which often fail when ingredients vary in size or defrosting state. Embodied AI refers to artificial intelligence systems embedded in physical bodies that perceive their environment through sensors and act upon it via actuators, enabling robots to interact dynamically with the physical world.