🤖AGIBOT's Six Robots Run Live Store Trial, Four Pass Performance Test
Robot retail just got a real performance review
TL;DR
AGIBOT ran six G2 robots through a 24-hour livestreamed store trial, and four passed the performance assessment. They handled coffee, restocking, cleaning and customer service across four rushes. The shared Genie Master framework coordinates tasks and learns from errors, a useful pattern for anyone building multi-agent systems.
Four of six AGIBOT robots passed a performance assessment after a 24-hour livestreamed store trial that ran from 10 a.m. November 2 to 10 a.m. November 3. Two failed to meet all the criteria, which is the more honest part of the story. The trial drew over two million viewers and closed out AGIBOT's AI WAVE showcase, which launched October 22. If you build agents or robotics software, the interesting part is Genie Master. It's an embodied AI agent framework that assigns work by each robot's capabilities and store conditions, then pulls execution results back for analysis and troubleshooting. Failed attempts go into long-term memory so similar situations get handled better next time. That's the same retry-and-remember loop many of us are hand-rolling for LLM agents. The robots logged 683 customer-service interactions, 369 product pickups and deliveries, 147 coffee orders, 126 breakfast servings, 277 shelf inspections and restocks, and 128 table cleanings. Evaluators scored them on time on duty, completion, customer feedback, response time, and handling of unexpected situations. AGIBOT says it will feed this operational data back into reliability and adaptability work and check which skills transfer across settings.

Key Points
The robots completed 683 customer-service interactions, 369 product pickups and deliveries, and 147 coffee orders across four operating periods, including the lunchtime rush and an overnight shift.
Keep reading — it's free
Enter your email to keep reading — plus the best of AI & tech, daily. Free, forever.
Already a member? Sign in
Comments
Be the first to comment