"By 2025, more than 70% of the manufacturing enterprises of the scale will basically achieve digital networking, and build more than 500 intelligent manufacturing demonstration factories that lead the development of the industry. The maturity level of intelligent manufacturing capabilities has been significantly improved." Not long ago, eight departments including the Ministry of Industry and Information Technology and the National Development and Reform Commission issued the "14th Five-Year Plan for the Development of Intelligent Manufacturing", proposing to promote the digital transformation, networked collaboration, and intelligent transformation of the manufacturing industry. and other measures.

As the main direction of building a strong manufacturing country, accelerating the development of intelligent manufacturing plays an important role in consolidating the foundation of the real economy, building a modern industrial system, and realizing new industrialization. At present, the development of intelligent manufacturing in my country has shifted from concept popularization and pilot demonstration to a new stage of system innovation and deepening application. However, compared with the requirements for high-quality development of the manufacturing industry, there are still many problems such as a high threshold for intelligence and weak innovation capabilities.

There is an urgent need for digital transformation of traditional manufacturing industries

"Intelligence is just a means and a tool. No matter how advanced intelligence is, it needs to be integrated into the manufacturing entity to be effective, otherwise intelligence will only be a castle in the air." According to Li Peigen, director of the National Intelligent Manufacturing Expert Committee and academician of the Chinese Academy of Engineering, the development of intelligent manufacturing must be settled. In manufacturing, and manufacturing is based on advanced technology and equipment.

Since the "13th Five-Year Plan", China's intelligent manufacturing development has made great progress. The market satisfaction rate of intelligent manufacturing equipment exceeds 50%; traditional industries are rejuvenated, and the proportion of investment in technological transformation in industrial investment has increased to 47.1%. However, with the increasing personalized needs of users and more intense brand competition, the uncertainty has increased significantly. For traditional manufacturing enterprises, accelerating digital transformation has become an urgent topic of the general trend.

In the process of promoting the intelligent transformation of the manufacturing industry, the integration of new-generation information technologies such as AI has brought new opportunities. Manufacturing enterprises use the gradually mature AI technology to reduce costs and increase efficiency on the basis of specific business scenarios, and realize short-term verification and application of AI models. This is not only the first step towards intelligent transformation and upgrading, but also a step into the intelligent economy. One of the keys to enjoying AI dividends on the track.

For example, in industrial manufacturing enterprises, the quality inspection process accounts for about 40% of the labor cost of the factory, and the quality inspection rate directly affects the growth of enterprise orders. At present, relevant companies have taken the lead in seeking AI technology for quality inspection and transformation, but the results are not satisfactory because there is no accumulation of relevant talents and technologies.

In addition, in the aspect of safety supervision, many companies hope to improve their automated supervision capabilities by combining AI technology, but they have not implemented them due to the difficulty of customization and high deployment costs.

For many manufacturing companies, common problems such as high accuracy requirements of AI models, large amounts of data required, and lack of experienced professionals have blocked the pace of leveraging AI to achieve intelligent transformation.

How to lower the threshold for enterprises to enter the first door of intelligent upgrade?

Seeking to cooperate with open source and open companies with leading AI technology capabilities may be the solution.

As a high-tech enterprise that provides solution design, system integration and other services to manufacturing enterprises, Zhuhai Huachuang Intelligent Technology Co., Ltd. (hereinafter referred to as "Huachuang Intelligent") is well versed in the pain points of quality inspection. During the inspection of the solutions, they found that the customization cost of many technology manufacturers is high, and the technology is difficult to meet expectations. Occasionally, Huachuang Intelligence learned that Baidu Flying Paddle Easy-DL zero-threshold AI development platform provides automatic modeling of multiple business scenarios, and also enables application developers without algorithmic foundation to successfully complete model training.

In the absence of professional AI algorithm engineers, Huachuang Intelligent took the real data collected on the quality inspection line through EasyDL for labeling and automatic model training with a try mentality. After efficient model iteration, the final accuracy was as high as 99% % of high-quality AI models are seamlessly integrated into third-party systems to realize real-time detection of assembly line product accessories.

After taking the "reassuring pill", Huachuang Intelligent helped gas stove manufacturers to build an assembly line accessories identification management system through Easy-DL, and completed the closed-loop judgment of all assembly line products, which can work 24 hours a day without interruption and complete the accessories for all products. Quality inspection, work efficiency is increased by 30%, and product quality inspection quality is increased by 80%.

In addition to production quality inspection, EasyDL has also penetrated into a series of scenarios such as sorting and packaging of manufacturing items, process optimization, and equipment maintenance, helping to reduce costs and increase efficiency while expanding more possibilities for business.

Jiangxi Ruishi Intelligent Equipment Co., Ltd. (hereinafter referred to as "Ruishi Intelligent") found in the investigation of traditional textile enterprises that the purchase of fabrics relies on manual search, which is slow and inaccurate, and the inspection of fabrics is also inefficient due to inconsistent standards. There are a wide variety of fabrics and they continue to increase, and a single detail will vary greatly. After trying some technical solutions, the company has been unable to solve the problems of slow model training and failed fabric quality inspection.

Under the recommendation of peers, Ruishi Intelligence first imported fabric pictures through the EasyDL platform, created hundreds of fabric labels, and trained a fabric recognition model. Unexpectedly, the training speed has been greatly improved, and the continuous introduction of new types of fabric pictures can also quickly Iterates and maintains over 99% accuracy. With confidence, Ruishi Intelligent has developed a complete intelligent system based on EasyDL. As long as you take a photo and upload it, you can obtain the different specifications, names, weights and other information corresponding to the fabric; at the same time, it can also accurately detect the number of fabric defects and mark them. According to the location, all participants in the industry chain and fabric testing information are circulated on the system in the form of "graphics and texts", and the evaluation is fair and the communication is efficient. "Easy to use, fast model training, and a model that can train well with a small amount of data are the three major features of EasyDL."

The digital transformation of manufacturing is a "required course" related to survival and development. Flying Paddle EasyDL not only helps manufacturers take the first step in the absence of AI technology and experience, but also solves the current problem of an accessory, a piece of fabric or a production line; in the long run, the technical application behind it is essentially It is the continuous exploration of many industrial enterprises from extensive production to intensive cultivation, further focusing on the main business, strengthening innovation, and gradually extending to the entire industrial chain of intelligent upgrades, which will have a great impact on the improvement of quality and efficiency of the entire industry.

At the critical stage of high-quality development of my country's economy, in the transition stage from product output to software and hardware output, a new generation of technologies represented by AI has gradually become one of the core driving forces of industrial transformation. Baidu, which owns the world's leading AI technology, etc. Enterprises are bringing new historic opportunities to China's manufacturing industry.