In "other people's schools", I can see the future of smart canteens...

After being away from campus for many years, I didn't expect that the cafeteria has undergone earth-shaking changes...

When you recall your student days, the first thing that comes to your mind is the "sadness" of staying up all night to finish homework before the final exam, the sweetness of campus romance, or the joy of skipping classes to play basketball with your roommates... But I wonder how many people are like me, thinking about my alma mater's cafeteria all the time, especially the meat pie on the breakfast menu at 2.5 yuan, which is definitely more intimidating than any alarm clock, urging me to get up early on time.

I believe that many people still have mixed feelings when they think of the school cafeteria again many years after leaving school. The bustling queue for food, the cafeteria auntie who always has her hands shaking when serving the dishes, and dark dishes such as fried tomatoes with bananas and orange slices, these little things that were extremely disgusting back then are unforgettable.

Recently, I wanted to take the opportunity to visit "other people's canteens" and reminisce about college life, but I didn't expect that the school canteen has undergone earth-shaking changes...

Obviously, the endless queues and the checkout aunties who could do mental arithmetic with zero mistakes are gone, replaced by self-service checkout smart terminals. Students only need to put the tray full of dishes in the checkout area, the total price will appear on the screen, and after swiping the card to pay, they can enjoy the food. The whole process can be completed in just a few seconds.

 A university in Beijing uses AI recognition settlement platform under Pushi Technology

Within a few minutes when the editor was amazed, four self-service checkout machines served nearly a hundred teachers and students. The flow of traffic from picking up meals to checkout was reasonably designed, and there was basically no queuing during the period.

Smart Canteen Blue Ocean Market

Faced with the surging wave of digitalization, all walks of life have actively invested in the work of transformation and upgrading, and the catering industry naturally cannot fall behind. Nowadays, in terms of takeout, the proficiency of Meituan and Ele.me platforms in the use of big data is obvious to all; in terms of dine-in, scanning code ordering and electronic payment have also developed very maturely. However, the digital transformation of the group catering industry represented by colleges and employee canteens started much later and has been tepid.

Different from the privatized nature of the social catering industry, the group catering industry has more advantages in the process of digital reform due to its large scale, abundant funds, and diversified demand. At the same time, in the post-epidemic era, the demand for intelligent and unmanned canteens is increasing as students return to school and companies resume work, which has also opened up a blue ocean market for smart canteens to some extent.

First, the number of diners in university and employee canteens is more concentrated and fixed than in social catering, and the volume of data provides a natural fertile ground for the implementation of technologies such as the Internet of Things and big data; secondly, canteens have stronger welfare and service natures, and when implementing intelligent solutions, managers mainly focus on actual pain points such as improving settlement efficiency, reducing congestion, and optimizing consumer experience, and will not pay too much attention to benefits and costs; in addition, with the rapid development of smart campus and smart park industries, canteens, as an important link in the ecological chain, their upgrades and renovations will further complete the intelligent scenarios within schools and enterprises and strengthen data management.

The smart canteen system can be divided into smart ordering, self-service settlement, smart meal collection and other solutions for different forms such as traditional ordering stalls, self-selected food stalls, and self-service food collection. Among them, self-service settlement for self-selected food stalls is the most widely used, which can not only improve the efficiency of meal collection settlement and thus reduce queue congestion, but also reduce costs by "replacing people with machines". The core of self-service settlement is the accuracy and timeliness of dish recognition.

How to achieve smart settlement?

Looking at the technical development path of dish identification products, in the early stage, it mainly relied on RFID chip identification technology, that is, a chip was implanted in each bowl and plate, and cooperated with the self-service checkout machine to identify the dishes in the bowl and plate and match the established price entered in the background. Although this move has high accuracy, the cost will continue to increase with the loss of the chip. At the same time, the bowls and plates for public dining need to be treated with high temperature, but the chip will fail to sense or be damaged if the temperature is too high.

In recent years, thanks to the rapid development of artificial intelligence technology, visual recognition has been widely used in scenarios such as face payment, video surveillance analysis, license plate recognition, and industrial defect detection. It has also successfully enabled the smart canteen industry and provided a new way to identify dishes. AI visual recognition settlement uses image recognition, edge computing and other technologies, relying on high-performance AI processors and large-scale databases. It can achieve rapid dish recognition without replacing the existing dishes and plates in the canteen, so the later operation and maintenance costs are greatly reduced, and there is no need to consider chip loss.

As one of the earliest artificial intelligence companies in China to engage in food recognition research, PuShi Technology focuses on visual recognition in the group catering field. It has developed a number of smart products for canteen scenarios, including AI recognition checkout counters based on image recognition technology, self-service checkout smart scales, dual-screen facial ordering machines, AI food collection cabinets, etc.

 The canteen of Guangdong Provincial Women and Children's Health Care Hospital adopts the dual-channel AI recognition checkout counter of Pushi Technology

朴食科技董事长陈晓鹏在接受物联网智库专访时介绍道，朴食科技AI识别结算台搭载华为昇腾AI处理器及工业级图像识别摄像头、依托公司自研算法优势，可在100毫秒内快速、准确地动态化识别菜品，从而保障结算流程的流畅与高效。值得一提的是，朴食科技的AI识别算法可在1秒内完成图像学习，即当结算台上出现未录入系统中的全新菜品时，机器可以在1秒内采集图像并完成建模，待工作人员在系统中输入价格后便可以直接在下一次结算中使用。正如陈晓鹏所说，朴食科技已经把教计算机认识菜品这个环节做到了极致。

 CSPC Pharmaceutical Group adopts Pushi Technology's dual-channel AI recognition settlement platform to support facial recognition payment

Group dining data drives the implementation of a health platform

There is no doubt that PuShi Technology is a leader in AI capabilities in the field of group catering. The company's smart canteen solutions have been maturely implemented in many projects, including the Changsha Sany Heavy Industry Canteen, the Guangdong Provincial Women and Children's Health Hospital Canteen, the Yantai Wanhua Global R&D Center Headquarters Canteen, Shijiazhuang Shijiazhuang Pharmaceutical Holding Group, and the Ocean University Cunjin College Canteen. It has empowered more than a thousand canteens, collected 270 million pictures of Chinese food, and formed a high-quality food picture training library, thereby continuously iterating and upgrading its self-developed AI algorithm.

With the implementation of the smart canteen solution, in addition to pictures of dishes, the system can also collect multiple types of dining information including consumers' eating habits, eating time, and payment methods for each meal. In an era of data-driven change, where will this "precious wealth" go?

First, it can provide data support for related research projects. For example, in the research process of topics such as "physical nutrition quality level of primary and secondary school students" and "eating habits of college students", the smart canteen system that communicates with the campus card can accurately screen the data of a given age group for reference; in the process of optimizing hospital diagnosis and treatment plans, the smart canteen system can track patients' daily diet and nutritional intake data to provide reference for doctors' diagnosis.

Second, it can be used to build a health and nutrition management platform. In recent years, consumers have paid more and more attention to personal health management. The health management platform built based on the smart canteen can automatically associate and record the dishes consumed by users at each meal, and conduct relevant data analysis to calculate the calorie and vitamin content, helping users to grasp the daily nutrition and calorie intake in real time.

Third, it can provide guarantee for the traceability of canteen dishes. That is, when there is a problem with the raw materials of a dish, the canteen manager can easily check which user has taken the dish.

PuShi Technology has been deeply involved in the smart canteen industry for many years and has independently developed multiple data modules including plate management, dish management, revenue data visualization, food chain transparency, and food traceability management, and provides personalized function customization for users on PC and mobile phones.

At the same time, PuShi Technology is gradually upgrading its strategic direction to health and nutrition management, providing data services to customers through the company's rich product matrix and health management platform. According to Chen Xiaopeng, chairman of PuShi Technology, the company has currently opened some AI capabilities to the Huawei developer ecosystem, and through cooperation with cash register software companies, the coverage scenarios will be expanded from canteens to supermarkets, fruit stores, etc., so as to collect more data.