A new automated shopping platform has been launched that uses far fewer cameras than other platforms.
In recent years, as the accuracy of computer vision has improved, a number of startups, such as Amazon Go, DeepMagic and Smartcart, have focused on creating automatic store checkout platforms. Now a new startup, Zippin, has launched an automated store.
Zippin founder Krishna Motukuri was inspired when he went to pick up a carton of milk one day, saying, “As soon as I stepped in the store, I saw the lines, and I knew there was no way I was going to go in for one item. That got me thinking, there’s gotta be a better way.” To demonstrate their automated shopping platform, the startup is launching an open beta prototype of the store in San Francisco’s SOMA neighbourhood.
Zippin uses AI, machine learning and visual cognition technology to allow shoppers to avoid checkout lines. Users first download the Zippin app and input their preferred payment method. The app assigns a personalised QR code to each user. This store ‘key’ is scanned as customers enter the shop. Once inside, overhead cameras follow each customers’ movements. More cameras, and shelf weight sensors, track products as they are picked up and/or put back on the shelves. The tracking system automatically registers which items have left the store, charging the customer and sending them a receipt.
Zippin claims that a store of around 1,000 square feet would need only about 15 cameras. This compares favourably to other automated stores, which use hundreds of camera arrays. The company estimates that the cost of converting a store that size would be around 25,000 USD. There would also be a monthly fee for using Zippin’s technology, based on square footage and sales volume. Unlike previous attempts to fully automate stores, Zippin sees this as a way to help stores drive up sales by serving more customers, rather than replacing traditional stores altogether.