AI and IoT: Two Technologies that are Shaping Staff-less Stores


20 September 2017

The Staff-less store is not a new thing, but since Amazon opened its AI-powered grocery store "Amazon Go" for internal employees in Seattle last December and announced its plan to build 2,000 grocery stores across the US, staff-less stores equipped with AI or IoT technologies are now getting more and more attention.


While the "Amazon Go" in Seattle is still not open to the public due to technical reasons, in China, the intelligent staff-less store has become widespread across the country and is favored by venture capital. The startup BingoBox has opened several staff-less convenience stores since June 2017; DeepBlue Technology released "QuiXmart" retail solution and introduced its staff-less "TakeGo" store during the "Artificial Intelligence Change Retail" Summit on June 25, 2017; and the leading Chinese e-commerce giant Alibaba, debuted its staff-less store "Tao Cafe" during the 2nd Taobao Maker Festival which was held in Hangzhou International Expo Center from July 8th to July 12th.


These staff-less stores all feature self-service, cashier-less payment, and range from grocery stores covering hundreds of square meters to convenience stores with only a few dozen square meters in floor area. According to the technology used to track what goods have been picked by customers, the staff-less store can be divided into "Machine Vision based in AI" and "RFID based in IoT".


Machine vision based AI stores (also known as "Computer Vision", the technology and methods used to provide image-based automatic inspection).

·         Examples: Amazon Go and TakeGo

·         The shopping process: registered customers check in at the entrance by mobile APP and confirm identity by palm prints or facial recognition, once they've finished shopping they can simply walk out the shop, the bill will be automatically charged through mobile payment. This kind of store uses cameras and sensors to watch customers and track what they have picked up, the technologies behind are a combination of machine vision (including facial, object, and human movement recognition) and deep learning.


RFID based IoT stores

·         Examples: BingoBox, Lawson, and 7-Eleven Signature. 

·         The shopping process: begins when registered customers check in at the entrance by mobile APP and confirm identity by facial recognition. After selecting the goods, customers place items on a scanner that recognizes the RFID tags all at once and proceed with mobile payment. 


Machine vision based technology has been mainly applied in the large grocery stores, and RFID based technology has been mostly applied in small convenience stores. But some stores have adopted both machine vision and RFID technologies at the same time, such as "Tao Cafe". No matter which type of staff-less store, facial recognition is important and must have the technology to identify customers. To compare these two types of staff-less stores, the cost of the RFID technology is relatively low from business point of view; but from a customer perspective, machine vision can provide a more simple and efficient shopping experience. Now is too early to judge which type of staff-less store will take the lead, but such technological innovations will undoubtedly bring new opportunities to the traditional retail industry.

About Marie Ma

Marie Ma is currently the general manager of Comba Telecom Network Systems Limited. Ms. Ma is responsible for overseeing the strategies and development of the new solutions and product marketing. She graduated from Tsinghua University with a master degree in Information & Communications Engineering in 2007 and a bachelor’s degree in Electrical Engineering &Automation in 2004. Ms. Ma has wide experience in product management, technical marketing and business development. She joined the Group in 2007.

More Articles