With the development of internet technology and e-commerce, China's physical retail industry has been increasingly impacted. In this context, commodity circulation channels have become more dispersed, consumer demands have become more diverse, and the need for instant consumption has grown stronger. Unmanned retail emerged in response. Since last year, many white-collar workers have noticed one or two snack-filled shelves suddenly appearing in their offices, even offering fresh food like boxed meals, salads, and yogurt, allowing consumers to purchase food without leaving the office, greatly satisfying instant consumption needs. As a new commodity circulation format, unmanned shelves can not only effectively reduce labor costs but also meet consumers' increasingly strong demand for instant consumption. This has led to frantic capital pursuit since the inception of unmanned shelves. According to incomplete statistics, in just over a year, unmanned shelf startups have accumulated financing exceeding 3 billion yuan. With the entry of internet giants like SF Express and Ele.me, the market prospects for unmanned shelves have been widely optimistic among industry insiders. At the same time, problems exposed during the rapid market expansion of unmanned shelf companies have gradually emerged. 1 Severe Inventory Loss Since the birth of unmanned shelves, the most questioned issue has been theft and inventory loss. Incidents of consumers taking goods without paying or underpaying are common, with industry-wide inventory loss rates generally as high as 30% or even higher. "In fact, unmanned shelf players are not unaware of the importance of anti-theft mechanisms, but in the first stage, they primarily consider market issues. The technology-free unmanned shelves are just transitional retail terminals for companies to quickly seize market share at this stage," a relevant person in charge from DeepBlue Technology told reporters. DeepBlue Technology is a leading runner in applying artificial intelligence to the retail industry in China and is currently a leading enterprise in the unmanned convenience store industry that uses AI technologies such as neural networks and machine vision to upgrade current retail. In response to the severe inventory loss in the unmanned shelf industry, DeepBlue has newly launched the Xiaolan ambient temperature shelves (semi-open and open shelves), targeting high-end office scenarios and park factory scenarios respectively. The newly launched shelves by DeepBlue Technology use the most advanced facial recognition technology. When someone enters the shelf recognition area, if not registered, a voice prompt will ask for registration; if already registered, "face scanning" (generating Face ID) to enter the shelf area will receive a welcome voice prompt, and can also intelligently recommend products based on previous consumption records. Customers can freely choose, and upon leaving, the system automatically recognizes products and completes settlement. This completely changes the cumbersome shopping mode of traditional unmanned shelves, requiring no phone throughout the process—just take and go. At the same time, it effectively reduces the occurrence of taking goods without payment and lowers loss. 2 Low Average Transaction Value Currently, the product differentiation of ambient temperature unmanned shelves on the market is very small, with products basically limited to packaged snacks under 10 yuan, and products priced above 15 yuan are rarely involved. "Although it is known that high-priced products have relatively higher gross margins, many unmanned shelf players still dare not try. Because in an open shopping scenario, high average transaction value means higher theft and loss rates, which may exceed their acceptable loss limits. To solve the low average transaction value problem, differentiation must be achieved through products like fruits, fresh produce, and cold drinks." To this end, DeepBlue Technology has specially designed and developed a Xiaolan freezer customized for office shelf companies, focusing on fresh food such as fruits, fresh produce, boxed meals, salads, and cold drinks. Combined with the previous two ambient temperature shelves, it can basically support the full-category product sales of office shelves, meeting the breakfast, lunch, and dinner needs of office workers as well as the need for snacks and energy replenishment during overtime. For unmanned shelf companies, providing fresh and cooked food products can on one hand form a differentiated advantage in the highly homogeneous competitive landscape, and on the other hand enhance consumer stickiness and increase customer loyalty. 3 Poor Shopping Experience Currently, the common product purchase process for office shelves is scanning a QR code to enter an APP or mini-program to select products and pay. It often happens that the purchased product cannot be found in the mini-program, requiring multiple scans of the product barcode to complete payment. The shopping process is cumbersome and complex, resulting in poor consumer experience. "The more we are in a technological era, the more we must use technology to drive industrial and industry transformation. Machine vision is the future direction of the unmanned retail industry. Products must be identifiable from a distance and in a non-contact manner to meet consumers' increasingly diverse consumption needs." "The future is the era of artificial intelligence, and the market belongs to enterprises that make good use of AI." The changes brought by technology are continuously affecting our lives. The future scenes that existed in science fiction movies are being realized one by one. Preliminary estimates suggest that the market space for office unmanned shelves has reached the level of tens of billions. The TakeGo "take and go" and super convenient shopping experience will lead the future trend of intelligent retail, and intelligent shelf upgrades will become the key to winning in the unmanned shelf 2.0 era. According to statistics, in 2017, the transaction volume of unmanned shelves and unmanned retail in China is expected to reach 10 billion yuan. In the next five years, unmanned retail stores will usher in a period of explosive growth. By 2022, the market transaction volume will approach 1 trillion yuan, and by 2022, the number of unmanned retail users will reach 240 million. With intensifying competition, the unmanned shelf industry is no longer just a competition for the number of points; refined operations and gradual profitability are the focus of competition for future survival. After the early stage of flocking in and accelerating point placement, office unmanned shelves will inevitably usher in intelligent upgrades. At the beginning of 2018, several major e-commerce giants entered the market, once again confirming the unprecedented huge market. Players who want to stand firm in market competition and strive to be among the few survivors must rely on the power of technology and artificial intelligence. **DeepBlue Technology is actively recruiting agents Scan the WeChat QR code for more details -END-