The Seattle store, known as Amazon Go, relies on cameras and sensors to track what shoppers remove from the shelves, and what they put back. Cash registers and checkout lines become superfluous – customers are billed after leaving the store using credit cards on file.
For grocers, the store’s opening heralds another potential disruption at the hands of the world’s largest online retailer, which bought high-end supermarket chain Whole Foods Market last year for $13.7 billion. Long lines can deter shoppers, so a company that figures out how to eradicate wait times will have an advantage.
Amazon did not discuss if or when it will add more Go locations, and reiterated it has no plans to add the technology to the larger and more complex Whole Foods stores.
The convenience-style store opened to Amazon employees on Dec. 5, 2016 in a test phase. At the time, Amazon said it expected members of the public could begin using the store in early 2017.
But there have been challenges, according to a person familiar with the matter. These included correctly identifying shoppers with similar body types, the person said. When children were brought into the store during the trial, they caused havoc by moving items to incorrect places, the person added.
Gianna Puerini, vice president of Amazon Go, said in an interview that the store worked very well throughout the test phase, thanks to four years of prior legwork.
“This technology didn’t exist,” Puerini said, walking through the Seattle store. “It was really advancing the state of the art of computer vision and machine learning.”
“If you look at these products, you can see they’re super similar,” she said of two near-identical Starbucks drinks next to each other on a shelf. One had light cream and the other had regular, and Amazon’s technology learned to tell them apart.