Autonomous retail cuts operational costs, personalizes shopping

Companies are starting to deploy popular AI technologies like RFID, facial recognition and sensor tracking to provide consumers with autonomous shopping experiences.

One of the most remarkable trends in artificial intelligence is the movement to autonomous retail through the use of unmanned operations. In January 2018, Amazon pioneered and popularized the concept of autonomous retail with their Amazon Go location in Seattle, which has subsequently expanded to almost two dozen locations in New York, Chicago, San Francisco and Seattle as of 2019. Is autonomous retail the shopping experience of the future?

The concept of autonomous retail is both simple and remarkable. Autonomous retail operations are brick-and-mortar stores that have employed an advanced and integrated product selection through check out technology that allows customers to simply walk in, select items off shelves and leave with minimal interaction with staff. These AI-enabled autonomous retail systems operate using similar types of technologies that are found in autonomous cars, such as computer vision, deep learning and sensor fusion. Technology can detect as well as track product movement in and out of the store and display selected items in a virtual cart. Sensors installed in the stores monitor the entire shopping process, checking the identities of each customer, recording all items each person picks from the store and, finally, automatically charging the total amount to the correct customer's account when they're ready to walk out. Typically, all that is required by shoppers is the use of a mobile app to check in and complete payment. This retail approach aims to cut operational costs associated with hiring human resources such as cashiers or customer support, while at the same time hoping to improve the shopping experience.

Key players in the autonomous retail market

While autonomous retail is still a novel concept with scarce examples available to most consumers, technology companies and retailers have been working toward the goal of frictionless retail for many decades. Prominent companies, such as Metro AG and IBM, have tried to implement this concept using RFID auto-checkouts, and other companies have experimented with QR code-based systems. However, it takes the creation of a completely integrated system from inventory to selection to check out and payment to make this concept work seamlessly. While Amazon has made waves with their well-publicized system, others are working quickly in their wake to make autonomous retail a ubiquitous reality.

Brick-and-mortar retail stores continue to experience disruption caused by the emergence and rapid growth of e-commerce, on-demand delivery and megastore dominance in terms of convenience, savings and choice. As a result, retailers are finding themselves having to adapt by focusing on more niche markets, delivering on unique selection, service or other aspects of convenience. IBM reported that over 80% of executives in both the retail and consumer products industries expect their companies to be using intelligent automation by 2021. Specifically, the report cited that the infusion of artificial intelligence into the automation process will enable smarter, more adaptive and more environmentally aware solutions that can not only make the shopping experience more frictionless but also make recommendations and further optimize retail processes.

China is heading the effort

In addition to the U.S., China is showing significant adoption and expansion in the world of autonomous retail. In China, mega e-commerce companies Alibaba and JD.Com are accelerating their adoption of unmanned retail stores. JD.Com has established a concept in grocery stores referred to as '7Fresh,' which has a range of autonomous retail capabilities including checkout and more detailed product information. Likewise, Alibaba has created a chain of digital grocery stores. In Hema stores, shoppers can scan and identify the origins of any product, place orders for food at the in-store restaurant, get relevant recipes, as well as receive personalized suggestions on products. Alibaba launched a cashier-free cafe back in 2017 called Tao Café, which requires shoppers to scan their phones or face to enter, and AI-powered cameras and shelf sensors track customers and products.

Other Chinese companies, such as BingoBox, have released entirely self-contained autonomous shopping experiences built into shipping containers that do not require any human interaction. With strategic strength in facial recognition and object recognition technology as well as a very advanced mobile-based payment infrastructure, China's embrace of autonomous retail shouldn't come as a surprise.

Not to be outdone by Amazon and their Chinese counterparts, retailers across the world are rushing to adopt this autonomous retail store concept. Starbucks has already implemented technology that uses geofencing technology and utilizes GPS and RFID within a physical location to generate virtual boundaries. These boundaries are manipulated to trigger feedback to mobile devices when crossed -- when a customer walks into a geofenced Starbucks shop, a breakfast meal coupon may pop up on their phones, prompting them to visit the store. Moreover, Starbucks uses geofencing technology in its order-ahead service. Customers order their beverage online and then walk into that local Starbucks to pick up their orders, where the geofenced boundary allows autonomous check-out based on customer ID. While not fully autonomous retail per se, this indicates the foundations of technology that will become necessary for future autonomous retail operation.

Will AI make retail jobs disappear?

According to some analysts, Amazon Go stores are already proving to bring in 50% more revenue per customer than typical convenience stores, and at significantly less operating expenses. The economic impact of autonomous retail, increased convenience to shoppers and greater ability to put small-footprint stores in many locations is going to make autonomous retail an inevitability rather than a curious proof-of-concept.

Like most AI efforts, this brings great efficiencies and convenience to store owners, but further jeopardizes the future of retail labor and work. According to the U.S. Bureau of Labor Statistics, almost 5 million people in the U.S. are employed in the retail industry, earning an average of $11.70 an hour. The bureau expects that with existing trends, retail employment will show a 2% decline, netting more than 100,000 total job losses over the next ten years. But with the pace of e-commerce, on-demand delivery and now autonomous retail, this might be a conservative figure. Margins in retail are extremely thin, with typical retailers experiencing net profit margins of only a few percentage points. The increasing capability of technology combined with the increasing expense of human labor will no doubt accelerate the trend away from human-to-human retail and result in the net loss or displacement of human jobs.

Altogether, the retail experience is shifting away from general purpose goods sold to a general audience to specialized retail offering niche products and experiences. It is hard to speculate what the retail environment of the future will look like, but there's no doubt that, given the trends already in play, it's leaning toward autonomous shopping and minimal human interactions.

Dig Deeper on Enterprise applications of AI

Business Analytics
Data Management