Denmark’s largest retailer is testing humanoid robots inside a Føtex supermarket near Copenhagen to handle one of retail’s most grinding, repetitive tasks: restocking shelves.

The machines, developed by Danish start-up Hive Robots and built by China’s Unitree, have been training since June. For now, human operators guide every movement so the artificial intelligence can learn how to grip and place different products. The goal is for the robots to restock shelves independently by early 2027, working mainly at night after the store closes.

The trial is part of a broader partnership with Salling Group, which operates more than 2,000 stores across six European countries. Føtex estimates that workers currently spend between 40 and 50 percent of their time simply replenishing shelves. Retailers hope the robots will eventually absorb a share of that physical work, freeing employees to spend more time assisting customers.

“Grocery stores and supermarkets are designed to function optimally for customers moving around the store. This makes traditional industrial robots unsuitable for retail environments,” said Simon Kudsk, Operations Director at føtex. “With humanoid robots, we gain a robot with the same physical form factor as a human — allowing them to move naturally and safely around the store.”

In later stages, the robots could also check expiry dates, identify empty shelves and spot misplaced goods. They still work more slowly than people and require close supervision. Danish law would need to change before the machines could operate autonomously among shoppers; for now, the focus remains on night-time work in areas away from customers.

Hive Robots founder Jakob Sand has described the current stage of development as similar to where conversational AI stood two to three years ago, emphasizing that real-world hours inside stores are essential for improvement. The company’s longer-term vision envisions multiple humanoid robots supporting staff in stores, though the present pilot remains carefully limited and supervised.

Whether the overnight machines can ease the burden on supermarket workers—or simply shift the night shift into a new form of remote human oversight—will be measured in the months ahead as the robots continue learning the aisles after closing time.