Customer Experience · 17 November 2026
Agibot Tests Six G2 Robots in 24-Hour Convenience Store Demo
Shanghai-based Agibot is running a livestreamed 24-hour convenience store demonstration with six G2 robots handling tasks from stocking shelves to preparing food. The event runs from 2 to 3 November, China time, according to Retail Customer Experience citing Humanoids Daily.
What happened
Shanghai-based robotics company Agibot is running a livestreamed, 24-hour convenience store demonstration in which six G2 robots are given store tasks, from stocking shelves to preparing food. The event began on 2 November and is scheduled to end on 3 November, China time, according to reporting by Retail Customer Experience, which draws on a Humanoids Daily report.
The robots are assigned to greet customers, make coffee, serve hot food, organise merchandise and collect rubbish. The challenge is built around the rhythm of a real trading day, covering lunch and evening rushes, overnight operations and breakfast preparation, so that performance can be observed across different demands rather than in a single showcase moment.
The reporting also flags an important limit. The livestream shows the robots working around customers, merchandise and service counters, but it does not establish that each unit can operate independently, without interruption or without human assistance.
Why it matters
For operators, the significance lies less in the hardware than in the question it raises for service design: can a physical, customer-facing environment be handed to autonomous systems across the full shape of a day? A convenience store is a demanding test, with frequent transactions, tight space, shifting demand and a mix of routine restocking and direct customer interaction. Demonstrations like this suggest robotic capability is moving from single, narrow functions towards multi-task roles in service settings.
The caveat matters just as much. Because the livestream does not confirm independent, uninterrupted operation, leaders should read the event as evidence of range rather than proof of reliability. The gap between a supervised demonstration and a dependable service model is where much of the customer experience risk sits, particularly when a delayed coffee order or a stalled hot-food station is visible to the shopper at the counter.
By the numbers
- 6 G2 robots assigned tasks in the demonstration
- 24-hour demonstration window covering lunch, evening, overnight and breakfast periods
- 2 November start date, with the event scheduled to end on 3 November (China time)
The Renascence take
Most coverage will dwell on the spectacle of six robots making coffee and clearing bins. The more useful question is which operating moments were chosen, because the lunch rush, the overnight shift and breakfast preparation are where service promises are most likely to be tested.
The real signal here is not whether a robot can pour a coffee; it is whether a store can keep its service promise when no member of staff is watching the counter. Customers judge a service by its worst recent moment, so one dropped order during the evening peak will shape perception far more than a flawless morning. Operators experimenting with robot-led zones should build visible fallback routes, clear human escalation points and service-level measures tracked per shift rather than per demonstration. Treat this livestream as a stress test of the handover between machine and person, not as a finished service model.
Sources
This briefing was written by our Newsdesk, synthesising reporting from the outlets below. Follow the links for the original coverage.
FAQ
Questions we get on this topic
More in Customer Experience
Stay ahead of CX
Get the signal, not the noise.
The stories shaping customer experience — plus the Journal and Experience Loom — in your inbox.


