Amazon’s Checkout-Free Technology Expands, Featuring Advanced AI
Amazon’s Just Walk Out technology, which enables checkout-free shopping, is now operational in more than 200 locations worldwide. The expansion includes stores on college campuses, as well as in stadiums, airports, and hospitals.
Ritu Subash, Head of Professional Services – Just Walk Out Technology, shared her excitement on LinkedIn, stating, “Incredibly proud to lead the team that deployed Just Walk Out at 200+ stores! If you haven’t already, experience the magic of JWO at a store near you.”
University Adoption and AI Advancements
Last month, Amazon announced the addition of several college campuses to the Just Walk Out system. These included Emory University, the University of Pittsburgh, the University of Maine, and the University of Virginia, bringing the total number of campus locations to over 30 globally.
In a significant advancement, Amazon recently introduced a new multi-modal foundation model that increases the accuracy of the Just Walk Out system. This new model uses transformer-based machine learning methodologies, similar to those used in generative AI applications, but adapts them for physical retail environments.
Jon Jenkins, Vice President of Just Walk Out technology at AWS Applications, explained how the new system functions, saying, “We accomplish this by analysing data from cameras and sensors throughout the store simultaneously, instead of looking at which items shoppers pick up and put back in a linear sequence.”
Jenkins further noted the benefits for both retailers and shoppers, stating, “For retailers, the new AI system makes Just Walk Out faster, easier to deploy, and more efficient. For shoppers, this means worry free shopping at even more third-party checkout-free stores worldwide.”
How Just Walk Out Technology Works
Just Walk Out technology uses a combination of cameras, weight sensors, and AI to allow customers to purchase items without going through a traditional checkout process. This technology, which launched in 2018, was initially built using machine learning to identify which items shoppers selected. Previously, the AI system assessed shopper behavior sequentially, analyzing movement, item selection, and quantities in a step-by-step process.
However, this sequential approach could sometimes cause delays or require manual retraining, especially in complex scenarios. These might include obstructed camera views due to lighting issues or other shoppers.
Improvements with the New AI System
Jenkins emphasized the advancements of the new model: “The new Just Walk Out multi-modal foundation model for physical stores is a significant advancement in the evolution of checkout-free shopping.”
“It increases the accuracy of Just Walk Out technology even in complex shopping scenarios with variables such as camera obstructions, lighting conditions, and the behaviour of other shoppers, while allowing us to simplify the system.”
“The new Just Walk Out AI system is able to achieve higher levels of accuracy by analysing all sensor data simultaneously, rather than sequentially. It looks at multiple inputs – cameras, weight sensors, and other data – and prioritises what’s most important to accurately determine the variety and quantity of items selected.”
He continued, “It also uses continuous self-learning and transformer technology, a type of neural network architecture that transforms inputs (sensor data, in the case of Just Walk Out) into outputs (receipts for checkout-free shopping).”
In one example of its enhanced functionality, Jenkins described a shopper browsing multiple varieties of yogurt. As they reach for different items, another customer could reach for the same item, or the freezer door may fog up and obscure the camera’s view. In these instances, which could have once caused delays to determine item selection, the new AI system rapidly processes the information from various sensor inputs. These include data from weight sensors on shelves, learns continuously from all inputs, and then prioritizes the most important information to accurately identify each shopper’s selections.
This reduces receipt delays and simplifies the technology’s deployment for retailers.
The self-learning capabilities of the new AI model also minimize instances of model retraining, which were sometimes necessary in unfamiliar shopping scenarios. To enable continuous learning, the system relies on a 3D map of the store, which contains the layout of fixtures such as shelves and freezers, and an image catalog of store merchandise. The combination allows it to accurately identify every product.
Even with store remerchandising, the system can recognize shoppers’ behavior.
Watkins highlights the user experience improvements, stating, “The improvements to our AI system are so seamless that you will continue to enjoy the same contactless checkout-free shopping experience you’ve come to expect at Just Walk Out stores, all while protecting your privacy.”
“Upon entering a Just Walk Out store you present a form of payment at the entry gate. The system immediately associates you with that payment method and begins mapping your journey around the store, adding or subtracting items from your virtual cart as you pick them up or put them down.”
“Just Walk Out does not collect any biometric information – the system only tracks how you interact with the products and fixtures (such as shelves or fridges), correctly identifying the products and quantities you leave the store with.”
Currently, Just Walk Out is in over 170 third-party locations throughout airports, stadiums, universities, and hospitals in the US, UK, Australia, and Canada.
Amazon plans to launch more Just Walk Out stores in 2024 than in any previous year, more than doubling the number of third-party stores with the technology this year. Watkins concluded, “As we scale, the system will continue to learn from everyday shopping scenarios and raise the bar for accuracy and convenience, delivering the benefits of AI to retailers and customers around the world.”
