Automating Human-Less Retail: The Alibaba Group Drives Face Recognition Innovation in APAC

The Alibaba Group is currently assessing face recognition technology as a means to unlock delivery drop boxes for its customers. The endeavor is also supported by Cainiao Network Technology, one of Alibaba Group’s primary logistics affiliates, as well as other biometrically-focused companies from the Ant Financial Group (find further information in the ABI Insight, “EyeVerify Gets Spoof Detection Upgrade and Extends Reach with Alibaba Acquisition”). Ensuring additional protection (not to mention more robust biometric templates), the currently tested pilot system would cross-reference users’ facial characteristics with the last template stored in the governmental registration security databases. The vision for the technology is to have the verification occur using Alibaba’s already existing biometric database of half a billion registered users. Should the plan proceed as expected within the next few years, then the company will be able to boast management of the largest commercial biometric database for face recognition worldwide.

The Alibaba Group is not the first to invest heavily in face biometrics. The payment giant’s decision to move away from voice biometric payments as a primary authentication modality and give more focus to face recognition has been in the works for some time now. Mastercard’s “Selfie Pay” has trended impressively across the security news during the last year. Even more recently, Apple’s decision to move away from its tried-and-tested fingerprint touch ID for the iPhone X and into a new era for Apple smartphone products using face ID is even more evidence of the market turning toward face identification.

A Cainiao Network Technology paper released in late 2016 states that, by the year 2020, there will be approximately 154 packages delivered in China alone. At the time the report was published, Cainiao claimed to process approximately 70% all of packages in the region with an average of 42 million packages every single day. This market is, of course, ripe for the new wave of Internet tracking and authentication technologies and it seems that China will be one of the first countries to push face recognition for deliveries mainstream. At the time of writing (September 2017), the Alibaba Group is showing outstanding performance in the stock market (BABA in US: NYSE), reaching an all-time high of US$180.40, up from US$101.45 a year ago, with an impressive 80% growth rate, increasing investor support in the company’s and its affiliates’ endeavors.

Alibaba’s plans, however, do not stop there. A new concept revolving around face recognition in the booming retail market in APAC has already been put to the test. The company has created Tao Café, a cashier-less café, which opened its doors to the public in July 2017. The establishment has no employees staffing the usual posts and everything from customer ID to payment and product delivery is automated via the use of face verification. The system only allows users who have successfully registered with Alipay (Alibaba’s mobile wallet) or WeChat (social media app), who are then prompted to scan a QR code to enter the establishment.

After placing an order, the system then verifies incoming registered customers and automatically subtracts funds from their Alipay or WeChat account. An ingenious process is also being used to let customers know when their order is ready by continuously scanning all customers’ faces and matching them with their order. A video explains how this concept works. Describing the aforementioned procedure, a company employee states that “in the whole process, there is no human interaction.” Although the choice of words definitely could have been different, is the “lack of human interaction” in retail actually a selling point? The Tao Café concept certainly coincides with biometric technological advances, a fervent desire for “personalization” in retail and shopping strategies, and recent developments from the machine learning and automation front. While everything seems to work as expected in the (quite possibly) staged experiment, face recognition could fail in many areas: continuously monitoring multiple customers, moving faces in the waiting areas, partial coverage from hair, facewear, or clothing, etc.

This development and other biometric, socioeconomic recognition, and beacon technologies deployed in the physical retail market, have been labeled “smart” shopping and are the next evolutionary step for the retail market. ABI Research hypothesizes that the very next steps could be the addition of virtual assistants to artificially fill the void of the “human-less” experience, the expansion of the concept in other verticals (primarily location retail, entertainment, and casinos), and, in the long-term, the transition of the technology from face to long-range iris recognition