TEL AVIV – WalkOut, the developer of an autonomous end-to-end checkout platform, is introducing a device that can be retrofitted onto any shopping cart, utilizing AI, machine learning, edge computing, and high-precision cameras to identify each item placed into or removed from the cart.   

Instead of utilizing weight sensors that require legal-for-trade certification, as well as constant recalibration, WalkOut uses machine learning and computer vision technology. The mountable cart device’s multiple high-precision cameras, which utilize many different proprietary algorithms to accurately recognize products, identify the items as shoppers load or remove them with 99.2% accuracy, according to the company.  

Through a large touchscreen, the device also communicates with the customer to offer personalized recommendations, store navigation, supplementary product information, and relevant promotions based on the shopper’s history and location in the store.  

“Unlike our competitors, we’re utilizing only highly accurate machine-vision technology with an over 99 percent accuracy rate achieved after lengthy rounds of development and testing,” said Assaf Gedalia, CEO and CoFounder of WalkOut. “Not only do our retrofit smart carts mount on a store’s existing carts, but items are registered in real-time while no final checkout validation is required—so customers can really shop as they please.” 

For more technology and equipment news, check out our operations page.