The unattended or unmanned store system as a black technology is sought after in the marketplace, and the key is to support the shopping experience through AI. Currently most unattended stores use the RFID technology. But due to factors such as distance and frequency interference, this technology can’t ensure a 100% recognition rate. Besides, cost will be an issue if the RFID tags are used on a long-term basis, and therefore it has hampered the commercialization of unattended stores.
YI Tunnel developed smart shelves using the semantic action recognition technology, to judge consumer’s shopping behavior, and to recognize and obtain the grocery bill for checkout. This globally leading technology will completely replace RFID, to help retailers develop their unattended store business.
Composition of the Solution
●Smart shelves based on the image recognition and semantic action recognition technologies
●The “Paying with Your Face” access control system that can replace mobile payment
●The facial recognition system for members
●The AI-based anti-theft system, which recognizes theft using an algorithm and gives alert in advance
Features of the Super YI
●Smart shelves, which judge consumer’s shopping behavior, and recognize and obtain the grocery bill for checkout based on the semantic action recognition and image recognition technologies
●“Paying with Your Face”, to replace mobile payment and really provide a “take and go” experience
● The machine now is already able to recognize 10000 SKUs, and the number will increase to 100000 by 2018
●Apart from standard items, it can also recognize nonstandard items such as vegetables, fruits, clothes, and dishes
●Extending equipment’s service life: based on the deep learning technology, the machine has the ability to optimize and update it self.It can keep learning after deployment, and the efficiency and accuracy can continue to increase
●Reducing costs for sellers: The breakthrough technology can replace RFID tags/barcodes/POS machines as well as cashiers
●YI Tunnel provides an extremely competitive price compared to other AI vendors—at only 1/50 of the global giant’s cost
After a consumer scans his or her face to enter the store and pick any items he or she wants to buy, the system will automatically generate a grocery bill. When he or she leaves, the system will be able to recognize his or her face and automatically complete the checkout process, so that he or she can really “take and leave”.
Customer’s Success Story:CCOOP, China’s Amazon Go
About the customer: CCOOP, a company of Hainan Airlines, is one of the models of CCOOP Group Co., Ltd. for physical stores. Through franchising, it covers community supermarkets and convenience stores from 100 square meters to 1000 square meters in both urban and rural areas across the country, to quickly expand its physical distribution network, to combine online and offline operations together, and to ensure the efficient circulation of merchandises across urban and rural areas.
About the project: The CCOOP smart and unattended convenience stores (China’s Amazon Go) are novel convenience stores driven by AI technologies such as the semantic action recognition technology and the facial recognition technology. They can dramatically improve consumer experience, lower OPEX, quickly scale up, and produce a brand effect, so as to solve the bottleneck of traditional convenience stores. It was the first time in the world to apply AI technologies in the commercialization of a retail business.
●The world’s first commercial unattended store based on full-visual recognition AI technology (known as China’s Amazon Go)
●Super YI, enabling you to really “take and go”: the solution for full-visual unmanned stores—unattended and RFID-less
●Only 1/50 of the global giant’s cost
●The deployment of the first store is fast, taking only 3 months from data collection to project completion, and it takes just 1 week to copy the system to other stores later
The Services and Technologies Provided by YI Tunnel
●Sorting out the user experience during checkout
●Gathering SKU data and training on machine learning
●The design and production of smart shelves
●The semantic action recognition technology
●The development of the facial recognition system for members
● The development of system interfaces, such as the interface with third-party payment systems as well as the interface with the cloud-workbench system