Amazon Go Shows that the Future is Present

Amazon opened its first automated convenience store, Amazon Go, in Seattle this month. In an attempt to develop a shopping experience free of checkout lines, the company opened a store where customers do not check out with a cashier upon exiting the store. Instead, items taken from the store are automatically billed to the shopper’s Amazon account.

A New York Times report described the experience like entering a subway station. To enter the store, customers must download the Amazon Go app on their phones and then use it to scan themselves through the entrance kiosks. Shoppers then simply choose items off the shelves and place them into bags (no need for shopping carts). After walking out of the store, customers will see a receipt for all items “purchased” show up on their phones.

Amazon does not freely discuss the technology supporting Amazon Go. In general terms, once users scan into the store using the app, it kicks into high gear in the subtlest way. A high power surveillance system watches each shelved item, and makes note of which shopper picks up an item and bags it. Reversely, the system can remove items from a customer’s inventory when that individual returns an item to the shelf. Amazon’s store completes these tasks without attaching any sort of computer chip to each item in the store.

Bloomberg reported complications with this technology in 2017. Issues pertaining to malfunctions with the store was too crowded contributed to the store’s opening being pushed back roughly one year. This long development duration allowed Amazon engineers to increase the sensors’ precision. Recently, engineers are working to help the store detect occurrences like children eating products straight off the shelf. Naturally, there are still outstanding issues like which person to charge in a couple shopping together.

One of my initial questions when first hearing about Amazon Go was the reasoning behind investing so much research and capital into the store. Amazon’s official reasoning is because they wanted to answer the question, “Could we push the boundaries of computer vision and machine learning to create a store where customers could simply take what they want and go?” Does this necessarily indicate the beginning of a new leap forward for automation?

Currently, Amazon has no public plans to expand this enterprise. Furthermore, no talk has surfaced concerning expanding the store’s technology to Amazon’s other subsidiaries such as Whole Foods. For now, it seems, knowing it is possible to open an automated store is enough of a progress leap.

A related question this store raises is whether automation is now closer to minimizing the need for human labor. Amazon posits that this is not a concern. Humans do work at Amazon Go, just not as cashiers. Instead, attendants roam the store re-stocking shelves and advising shoppers on what to purchase. Furthermore, chefs work in the store preparing meals and ingredients. Amazon appears willing to keep job opportunities present, indicating the store is not meant to cut down on the company’s payroll, but reform the customer’s experience.

Critics of automation will likely note that if Amazon wanted to completely eliminate human employees from this store, it would not be difficult. Machines already exist that can prepare food and re-stock shelves. To me, this indicates Amazon has no intentions to test the concept of a store completely manned by robots. Instead, Amazon Go’s purpose is simply to add ease to the everyday shopper’s experience. Bypassing any sort of checkout line and, therefore, saving time is a facet of shopping traditional stores cannot provide. This is more so the lure of Amazon Go than the novel fact that it will never accept a bag-boy’s application.

Edited by Cheyenne Abrams.