Robot-Staffed Shop in Hefei Takes Around 100 Orders a Day

About 100 orders a day. That is what a small outlet in Hefei's Hecai 1972 creative park achieved during the October National Day holiday, with only robots working behind the counter. The figure is modest for an ordinary coffee shop or store, but in robotics it says more than a dozen videos of dancing humanoids. A robot that takes money and hands over goods to real people every day faces a very different test from a robot on an exhibition stand.

What we know about the shop

Jiqizhixin's report describes the situation this way: over the past year, humanoid robots have learned to run, dance and fight, and have begun appearing in factories, warehouses and shopping malls. But for the companies building them, the publication notes, a polished demo and daily work with customers remain two separate challenges. The Hefei shop is trying to close that gap.

Based on available information, the picture looks like this:

• city: Hefei, Anhui province;
• venue: the Hecai 1972 creative park;
• period: the National Day holiday, early October 2026;
• format: a small outlet where only robots act as operators;
• result: about 100 orders a day.

The robot models, product range, average order value and level of remote supervision are not disclosed in the available portion of the report. It is too early to draw conclusions about them.

Why the number matters more than a flashy demo

An art park during the holidays draws a large crowd, and curiosity about the robots also boosts sales. Many people come to look, not to buy. That is why 100 orders a day cannot be carried over directly to an ordinary weekday. Still, as a test of the technology, this kind of traffic is useful: it creates unpredictable load, queues, unusual questions and payment errors. On stage, a robot runs a pre-rehearsed script. Behind a counter, the script breaks every few minutes.

The scale of the deployment also deserves attention. This is not a factory or a logistics hub, where protocols and floor markings protect the robot. It is retail with live customers: children, elderly people, tourists with phones. An environment like this quickly exposes a system's weak points: speech recognition in noise, speed of service, safety around people, and the reliability of manipulators across hundreds of repeated operations.

What is still missing to draw conclusions

The biggest gap is the economics. It is unknown how much it costs to run such an outlet, how often a human has to step in, and what share of revenue goes to maintenance and repairs. Without these data, it is hard to say whether the shop pays for itself or serves as a technology showcase for the manufacturer. In China, where regional authorities and parks readily support robotics projects, the second option is quite likely.

What to watch next

For anyone following the market or planning retail pilots with robots, a few benchmarks are worth keeping in mind. First: whether the result repeats after the holidays end and the novelty effect fades. Second: the share of orders completed without operator intervention. Third: the time it takes to serve one customer compared with a human. Fourth: the hourly cost of running the system, including maintenance. If the Hefei outlet publishes even some of these metrics, it will become a useful benchmark for the whole industry. If not, it will be one more eye-catching case from which it is hard to judge how mature the technology really is.