The AI Cooking Research Restaurant at Beijing’s Haidian Canteen has introduced full-process artificial intelligence management across its daily foodservice operations.
The system, which covers ingredient preparation, automated cooking, portion weighing, meal pickup, and real-time nutritional calculations, represents a significant integration of digital technology and commercial catering.
Developed in collaboration with a research team from the College of Food Science and Nutritional Engineering at China Agricultural University (CAU), the smart facility serves as both an active community dining hall and a testing ground for automated kitchen appliances and lab-verified nutritional database development.
Automated Back of House Kitchen Operations
The integration of intelligent kitchen machinery has modified traditional commercial kitchen workflows. Operating with a suite of automated cooking appliances, the kitchen can handle a wide variety of preparation methods, including steaming, boiling, roasting, deep-frying, and stir-frying.
To standardise food quality at scale, professional chefs program the cooking temperature parameters, ingredient ratios, and specific cooking durations directly into the automated devices. This automated system ensures consistent taste, appearance, and portion sizes across the entire menu.
The transition to automated cooking has changed back-of-house staffing requirements:
Traditional Labour Metrics: To prepare the same volume of daily meals, a conventional commercial kitchen would typically require more than ten chefs working throughout the morning.
AI-Enabled Efficiency: The automated Haidian Canteen requires only one head chef and two assistants to prepare and serve hot meals to 500 diners within a single hour.
To date, the automated kitchen system is capable of producing approximately 70 to 80 distinct types of pastries alongside an extensive selection of home-style Chinese dishes.
Smart Inventory Control and Food Waste Reduction
The application of real-time operational data has supported food waste mitigation and simplified inventory management.
A central digital dashboard in the back kitchen continuously monitors key performance indicators:
Sales and Usage: Tracking active dish sales and raw material depletion rates.
Inventory Reminders: Triggering automated reorder alerts for kitchen staff when ingredient stocks run low.
Customer Flow: Measuring hourly footfall to help management forecast peak demand periods.
According to restaurant manager Pan Qiang, the combination of smart kitchen tracking and a pay-by-weight, on-demand serving model has changed customer dining habits. By prompting diners to take only the portions they intend to eat, the canteen has reduced kitchen food waste by more than 90 per cent compared to traditional commercial dining facilities.
Visualised Nutritional Dining and Smart Chips
The canteen features a visualised nutritional dining system that links food selection directly to clinical health data. Each dining plate is embedded with an individual smart chip. Once a diner places their selected food items on the countertop plate, a terminal screen instantly displays the exact dish weight, unit price, and certified nutritional values:
🔥 Calories – Total energy delivery per portion.
💪 Protein – Amino acid metrics for muscle and cellular health.
🥑 Fat – Saturated and unsaturated lipid content.
🌾 Carbohydrates – Available carbohydrates and dietary fibres.
🧂 Sodium – Salt density monitoring to support cardiovascular health.
All nutritional statistics displayed on the terminals are analysed and verified in the laboratories of the China Agricultural University partnering research team. Currently, the project partners have finalised nutritional standardisation archives for 500 individual dishes, with a target to expand the standardised recipe repository to 1,000 entries by the end of 2026.
Future Targeted Nutrition Research
Launched on 22 December 2025, the smart canteen has maintained a stable operational volume, serving approximately 500 diners daily.
According to Professor Hu Xiaosong of the College of Food Science and Nutritional Engineering at China Agricultural University, the research team intends to further upgrade the facility's core operating software and hardware.
The next phase of the project's development will focus on targeted functional nutrition, designing specific dietary solutions to support adolescent eyesight, alleviate sleeping disorders, and assist diners managing chronic metabolic conditions such as hypertension, hyperglycaemia, and hyperlipidaemia. By embedding these target metrics into automated recipe formulations, the partners aim to demonstrate how smart commercial catering can actively support public health management.
.jpg)
