Amazon recently opened its revolutionary, no-check out convenience/grocery store, Amazon Go, at its headquarters in Seattle, WA. The store, announced in December 2016, opened to the public in January 2018 after beta testing for a year with employees.
We don’t know how the concept of a no-check out store will change the face of retail in the future. But for now, it’s truly marketing at the intelligent digital edge, the place where people and devices (and groceries) meet.
What is Amazon Go?
The Amazon Go concept is that customers can walk in, pick up items off the shelf and “Just Walk Out.” There are no baskets, no registers and no cashiers. But there are plenty of employees: shelf stockers, ID checkers in the wine and beer section and chefs making sandwiches and grab-and-go meals.
Here’s how it works. Customers download the Amazon Go app, which is linked to the customer’s Amazon account and the associated credit card. As they enter the 1800 square foot store, they scan their smartphone code at one of several glass security gates that are futuristic versions of subway turnstiles.
Shoppers then pick up the items they want. Once selected, the system adds the items to a virtual cart. If a shopper puts an item back on the shelf, it’s automatically deleted from the virtual cart. As customers leave the store, the app automatically charges the account and Amazon sends an electronic receipt logging what was purchased, how much it was and how long the customer was in the store.
Dilip Kumar, Amazon Go vice president of technology said “People are rushed. They’re in a hurry. People don’t like waiting in lines.” He added that the store concept is “to be respectful of your time as a customer.” The idea is to create an “effortless experience for customers.”
How Does it Work?
Amazon isn’t saying how the system works other than it combines computer vision, machine learning, artificial intelligence deep learning algorithms and sensor fusion technologies, much like the technologies that guide self-driving cars. Kumar says the idea behind the technology is to “push the boundaries of computer vision and machine learning.”
While the AI and machine-learning technology to make Amazon Go work was four years in the making, the foundational technology that enables the intelligence at the digital edge is based on wired and wireless connectivity.
Walk into Amazon Go and look up. You’ll see a hundred black, boxy cameras covering the ceiling. These cameras do the computer vision work, detect motion and see what interactions customers have with the shelves. (Amazon did say the cameras include infrared sensors, but that it is not employing facial recognition technology.) The cameras provide seamless detection as the shoppers move from area to area in the store. The computer vision enables computers to process the information and determine which items were selected.
The sensor fusion technology is embedded in the store shelves. It detects motion and registers when a customer picks something up or puts it back on the shelf. Because the system knows everything by its weight, customers can’t “accidentally” pick up two of an item. The last part of the technology is the artificial intelligence or the machine-learning algorithms, which enables the computers to determine what the item is and to learn by continuously collecting and analyzing data.
What’s Next?
The entire concept of Amazon Go brings up some questions. Does Amazon plan to roll-out this technology to its recently acquired Whole Foods stores or to its current 13 brick-and-mortar bookstores? The company says there are no plans for a national rollout and it plans on using the no-register shopping technology only at this Amazon Go. There are no more Amazon Go stores planned at this point.
In addition, the data collection and tracking technology Amazon is using can raise some privacy concerns. The data Amazon is collecting from its customers will accumulate exponentially over time. How will Amazon use that data? What about Amazon Prime customers? Will Amazon Go purchases data be incorporated in regular Amazon shopping suggestions and recommendations? Every time an Amazon Go customer visits the store, the Go systems learns more about their shopping behavior and preferences. As a smart marketer, it’s a good guess that Amazon will use this data tailor products and services for its customers and offer them a more engaging, personalized shopping experience at the digital edge in the store and online.
Foundational Edge Technology
Engineering the right foundational and enabling technologies for a smart store like Amazon Go is more than artificial intelligence and machine learning. It’s the right connectivity environment that makes it all possible. For instance, it’s the wired infrastructure that connects and powers the cameras. It’s the wireless capabilities that enable the system to talk to your smartphone. It’s the IT framework that provides 99.9% uptime. It’s a partner that understands the intelligent edge and can help you realize your digital transformation to some really cool technologies. Learn more at BlackBox.com/Intelligent-Digital-Edge
