SandStar Officially Launches a Smart Ice Cream Kiosk Enjoying a Clear Advantage Over the Competition

A smart ice cream kiosk of SandStar was launched at a major shopping center recently. Various ice cream cones and bars are neatly displayed in the fridge, waiting to give customers a cool and refreshing experience during the hot summer days.


​​​​​​​​​​L​ike other Smart Kiosks from the company, the smart ice cream kiosk also provides a convenient “grab-and-go” shopping experience. Using smart kiosks to sell ice creams is a new experiment for retail by SandStar. It has also demonstrated the advantage of the company’s Smart Kiosks powered by dynamic computer vision recognition as the key technology in real-world applications.

Among all the “grab-and-go” smart kiosk solutions available in the marketplace today, the solution from SandStar is based on dynamic computer vision recognition. In comparison, all other solutions (e.g. RFID, weight sensor, and static vision recognition) have obvious limitations.

With a smart kiosk based on weight sensors, for example, limited by the operating principle of weight sensors, all the items must be placed in the fridge following the original display pattern. With ice creams of numerous types, once someone puts an ice cream back to a spot it does not originally belong to, this will seriously compromise its recognition, and the user experience will be very poor. Besides, the weight of items such as ice cream cones and bars is very close to each other, so it’s very difficult for them to be accurately recognized by weight sensors, which must be used in combination with other solutions to function normally.

Static vision recognition requires that merchandise is arranged strictly in the sequence of being low in the middle and high on the two sides, and each layer of the fridge has to leave enough space in terms of height so that the camera’s vision won’t be blocked. Generally speaking, 40 percent of the space in a solution based on static vision recognition is wasted, the number of items that can be sold is reduced, and a higher restocking frequency is needed.

In real-world applications for products like ice cream cones and bars, a solution based on RFID tags has to bear the very high cost of disposable tags and the labor cost, and the RFID tags are also very easy to be peeled off, causing a merchandise damage rate of over 10 percent.     

SandStar’s Smart Kiosk based on dynamic computer vision recognition recognizes not just the merchandise, but also how customers take them and put them back, so the recognition accuracy is dramatically increased. In actual applications, they are subject to fewer restrictions while being able to provide a solution that can really help retailers reduce costs and increase efficiency.

  • When you display merchandise, you just need to think about if the display looks good and reasonable, and don’t need to create a strict display matrix based on the height and weight of each item, and therefore restocking is simple and fast;
  • How a customer takes an item and puts it back doesn’t affect the recognition of the item, so the user experience is better;
  • The space within a fridge is 100 percent utilized, leading to the increased operational efficiency and a reduction of OPEX of at least 20 percent.

In addition, SandStar’s dynamic vision recognition solution can also monitor surrounding traffic, gather and analyze the data about consumer behaviors, and help work out a better location strategy, merchandise mix strategy and pricing strategy. Based on the feedback from customers in actual use, the merchandise mix strategy and pricing strategy optimized with the data collected by  SandStar’s Smart Kiosk can help to increase sales by at least 20 percent. 

(SandStar acts on Corporate Social Responsibility (CSR) initiatives, the solution is carefully designed to protect customer privacy. The camera involved only captures products and the shopping behavior by marking the key point the human skeleton. Additionally, since the system is based on self-learning, all data processing is done automatically by the system itself.)

Therefore, SandStar’s Smart Kiosk based on dynamic computer vision recognition has become the choice for three leading global beverage suppliers.

As a star product of SandStar, its Smart Kiosk has been widely used at locations such as gas stations, convenience stores, shopping centers, and office buildings. They can be used to sell a variety of merchandise including snack-beverage combinations, toys with wrappers, and everyday necessities.

Thanks to the evolution of eight generations over the past four years, this year the Smart Kiosk as a star product of SandStar will help the company open up this niche market internationally, to enable customers around the world to benefit from its state-of-the-art solution that really works and generates value in actual applications.

About SandStar​

SandStar is an artificial intelligence company that provides leading computer vision technology for the retail industry. The advanced retail solutions engage shoppers like never before with grab-and-go convenience, autonomous checkout, and big data analytics, serving businesses worldwide. Founded in 2016, SandStar now has over 220 engineers. The company provides Smart Kiosks, Pure Computer Vision Smart Shops, and Smart Store Analytics. Dual-headquartered in Charlotte, N.C., and Beijing, SandStar serves 20+ Fortune 500 companies and strives to build an AI retail ecosystem. SandStar is a qualified Intel® IoT Market Ready Solution (Intel® IMRS) partner and Microsoft Co-Sell partner, supported by a number of leading venture capital investors after four rounds of funding, including CP Group, one of the world's largest conglomerates. Check out more on

Source: SandStar