A 250-Seat Canteen Lets Robots Cook, Stir and Wash

The signal: MAGIC CHEF makes programmable cooking and self-cleaning robots part of a 250-seat canteen’s operating system, with robotic plating and other automation added through a staged deployment.

The Concept

MAGIC CHEF is a 250-seat weekday canteen inside Tokyo’s Telecom Center Building, but its core experiment sits behind the counter. TechMagic’s I-Robo 2 handles part of the hot-food production by automating stirring, heating and the washing of its cooking vessel after a dish is finished. Recipes are translated into machine settings: heating temperature and time, pot rotation speed and rotation direction can be adjusted for each menu item. That allows dishes such as fried rice and stir-fries to be reproduced through a programmable process rather than depending entirely on one experienced cook.

The restaurant does not present the robot as a standalone spectacle. It combines robotic equipment with conventional kitchen appliances and human operations run by SENDAN, a CSS Holdings group company with experience operating hotel and institutional dining. TechMagic says the system is being introduced in three stages across May–July, August–October and from November onward, allowing the team to stabilise operations before increasing the degree of automation. The plan includes M-Robo plating technology alongside I-Robo 2.

Customers still encounter a familiar canteen format. The initial menu is based on everyday dishes, with greater choice around mains, side dishes and rice quantity. Tokyo Teleport Center’s official restaurant listing also says MAGIC CHEF uses fully cashless post-meal payment, so guests can collect food before settling the bill electronically.

The important point is that the restaurant itself functions as a live operating environment for food robotics. Instead of demonstrating one machine at a trade show, TechMagic is using a high-capacity lunch service to test how robotic cooking, human staffing, plating, payment and menu customisation can work together under real daily demand.

What Is Actually New

Cooking robots already exist, so MAGIC CHEF is not interesting simply because a machine can stir a wok. The stronger innovation is the attempt to redesign an entire canteen workflow around programmable robotic tasks and then increase that automation in stages while the restaurant remains operational.

I-Robo 2 does more than perform one visible cooking motion. It controls heat, timing and pan movement according to the recipe, then automates the normally separate cleaning step for its cooking vessel. That turns several repetitive kitchen tasks into one programmable cycle. In operational terms, a recipe becomes partly transferable software: temperature, duration and movement can be reproduced without relying on the same experienced cook standing at the station every service.

The second important element is scale. MAGIC CHEF is not a six-seat robot demonstration or an unattended vending kiosk. It is a 250-seat workplace canteen designed around a concentrated lunch period. TechMagic and CSS are using that environment to learn where robotics genuinely removes labour and where people remain necessary.

The phased rollout is also significant. Cooking automation comes first; plating automation and other process changes are introduced later rather than claiming instant full autonomy. That makes the project closer to an operating-system experiment for institutional foodservice than to robot theatre.

What remains conventional is the menu and dining room, deliberately so. The innovation is backstage: converting parts of kitchen craft, cleaning and eventually plating into repeatable machine processes that can be measured, adjusted and potentially copied into other high-volume foodservice sites.

Why People Want This

For customers, the strongest benefit is not watching a robot. It is what automation could remove from a high-volume canteen experience: long waits, inconsistent preparation and the need to choose only from food cooked far in advance. MAGIC CHEF is explicitly built around serving freshly prepared dishes while shortening service time, which matters when hundreds of people arrive during a narrow lunch window.

Programmable cooking can also make customisation easier to support operationally. The restaurant is increasing choice across mains, sides and rice quantity, while TechMagic has discussed even more flexible serving models in the future. If the kitchen can absorb those variations without requiring a skilled cook to manually reproduce every process, customers gain choice without necessarily slowing the line.

Consistency is another practical attraction. The robot can repeat defined temperature, timing and movement settings for the same recipe, reducing some of the variation that comes with fatigue or changing staff. That does not guarantee better food, but it makes the process more controllable.

The fully cashless post-payment model adds a separate piece of friction reduction: customers collect their meal and settle electronically afterwards rather than making payment the first bottleneck in the service flow.

The more interesting consumer promise is therefore almost invisible. If the system works, the guest experiences a normal lunch that is fresher, faster and more configurable even though the labour architecture behind it is substantially different.

WBC Verdict

MAGIC CHEF clears the innovation threshold because the restaurant is being used to redesign kitchen labour, not merely decorate the customer experience with a robot. Turning heat, stirring, recipe timing and vessel washing into one programmable cycle is a meaningful operating change, especially in a 250-seat lunch environment. The staged approach also makes the experiment credible: TechMagic is not pretending the kitchen is already autonomous. The key test is whether robotic cooking and plating actually improve throughput and staffing economics at scale without narrowing the menu or creating new maintenance bottlenecks.