Digitizing? Intelligent? What is the status quo of smart manufacturing in Chinese enterprises-Smart Manufacturing Channel

Most Chinese enterprises have basically completed the mechanization stage. They are at the stage of automation and digitization, while intelligentization has just begun. Smart manufacturing is mainly concentrated at the beginning and end of the production process, and is rarely used in the production process. The main strategy for advancing intelligent manufacturing is to encourage the shortcomings of automation and digitalization, focusing on supporting intelligence, and focusing on supporting the optimization of production processes in core areas.

To engage in "manufacturing", one should be proficient in engineering issues such as machinery, technology, products, and automation; while engaging in "intelligence" should be proficient in engineering modeling, machine learning (based on online data), and intelligent system architecture development. This article discusses the status quo of smart manufacturing in various countries from the perspective of mechanization, automation, digitization, and intelligence, as well as China's due diligence.

Over the past 30 years, in the field of smart manufacturing, countries have evolved from numerical models and offline models that focused on smart design in the early days to online smart systems with the goal of optimizing smart manufacturing processes, moving from traditional industries to emerging industries. Europe and the United States have a relatively good technical foundation and it is relatively easy to develop high-end technologies; but in China, due to the weak technical foundation, it is not easy to achieve the same effect.

The application of computers in the manufacturing industry began in the 1950s and split into multi-level computer control in the 1980s. Among them, the secondary system has all the functions of today's intelligent manufacturing system.

In the development of the factory intelligent system, Cascade led the development of three sets of secondary systems (intelligent systems) for metal smelting electric furnaces, refining furnaces and continuous casting; Oregon Corporation of the United States solved a series of production process optimization problems through intelligent systems, such as Elimination of defective products, successfully developed a new generation of secondary system that combines microstructure models, intelligent self-learning and continuous upgrades. The defective products in the production process of hard and thin products have been optimized from defective products every day to half a year. There has never been the same defective product.

Even in the early days when the term “intelligent manufacturing” has not yet become popular worldwide, these companies are already doing data collection, engineering modeling, machine learning, and intelligent system architecture development for high-end manufacturing, and use intelligent software to produce the best production process. The optimal parameter combination is transferred to the basic automated execution.

vacuum coating machine,pvd coating machine,pvd vacuum machine,vacuum ion coating machine,multi-arc ion coating machine

China has introduced some intelligent systems (secondary systems) in the iron and steel industry since the 1990s; however, the West still imposes an embargo on China in terms of models (only available DLLs are provided), which has led to China’s better engineering modeling. weak. For the 8,000-ton equipment, the operator would not dare to operate when the forecast value exceeded 4,000 tons; there would be problems such as making mistakes when formulating production procedures. The author participated in the secondary system optimization project. During the period, using the model advantages of this team, the investment utilization rate of the intelligent production line invested by Nanjing Steel (American Technology) with an investment of more than 10 billion yuan was increased by about 70%. After entering the emerging industry, he developed a series of projects, such as BYD lithium battery project. At the beginning of the cooperation, BYD used the extremely difficult to model pole piece segmentation burr prediction model to strictly inspect the model level, requiring the model hit rate of 85%, and the BYD lithium battery project team achieved a 98% hit rate, which made the project a success. The second phase (burr warning) and the third phase of the project (knife gap measurement device and model forecast), and in response to the weak data collection status of China's manufacturing industry, successfully applied the extremely difficult soft measurement technology in the industry. During this time, the author also completed a series of other projects, such as electronic manufacturing projects in more than ten companies including Skyworth, TCL and Guangye.

What is the status quo of smart manufacturing in Chinese companies? At least in the core links of intelligent manufacturing (optimization of the production process), such as engineering modeling, machine learning, and intelligent system architecture development, the situation is not optimistic.

In the on-site intelligent manufacturing project, the company has completed the project’s engineering problem modeling and intelligent system development, as well as offline measurement of key tool parameters and online soft measurement. Although the data provided basically meets the digital manufacturing represented by the data kanban, it provides high Quality data is not easy. The cost of such an intelligent system supplier is very high. Enterprises do not have a clear penalty system for the problem of missing data on site, resulting in poor on-site data integrity. Small businesses often simply cannot collect the high-quality data they need.

For the time being, we will not make a comparison between information technology and digital manufacturing technology, nor talk about the gap in industrial core software. Only from the core areas of intelligent manufacturing-engineering modeling, machine learning and intelligent system architecture development, Chinese companies are relatively What is the status quo in Europe and the United States? Looking at the four development stages of mechanization, automation, digitization, and intelligence, which stage are Chinese companies currently in?

By examining the basic requirements of smart manufacturing, the basic status quo of smart manufacturing in various countries can be detected from the degree of satisfaction of related requirements. At present, the machine generation is one of the efforts, but this is only at the level of automation. In order to achieve the most optimized machine generation, it is first necessary to model engineering problems and engineering parameters, and then use the collected high-quality data to perform machine learning of the model; the subsequent models are deeply bound to the machinery and production status. Based on these, intelligent systems can be developed, and then real-time variable production parameters that have always been optimized are generated, which are handed over to the basic automation for execution. This optimized machine generation is precisely intelligent manufacturing. There are not many of these in China, and they mainly exist in some platform vendors.

A large number of optimizations and expansions based on European and American technology (such as the development of a new generation of secondary systems), due to the huge system (often millions of lines of source code) and complex logic, when developing Chinese projects, they are mainly based on existing Active program rewriting. Western intelligent systems usually have experienced decades of technological accumulation, and China needs to improve in this field.

vacuum coating machine,pvd coating machine,pvd vacuum machine,vacuum ion coating machine,multi-arc ion coating machine

When the production process is not optimized enough, first perform diagnostic analysis without automation; this can be considered at the mechanization level. Until the related issues are resolved, the automation issues are reviewed. All of this can be quickly diagnosed and overall confirmed by digitalization, that is, when high-quality data is collected on the spot, the data can be used to identify which link is the weak point. Now there is a kind of "IT+OT technology" (information technology + operation technology) on the market, which reduces all engineering issues to OT in the digital phase (although this is a bit too simplified for intelligent manufacturing). Through the data kanban, many problems can indeed be seen, but the user of the data kanban must have a sufficient understanding of the production process. If the production line only has digital manufacturing, but not intelligent manufacturing, it is necessary to have engineers on-site guidance, because it is difficult for operators to see many problems from the data, and it is even more difficult to find solutions based on complex data.

China encourages the shortcomings of automation and digitization in the promotion of intelligent manufacturing, but this will also lead to misunderstandings of intelligent manufacturing and cause greater losses. For example, when an intelligent production line is being built, because it is not clear what intelligent manufacturing is, only production lines with basic automation, MES and ERP will be built, which is mistaken for intelligent production lines. What this type of production line lacks is the core of intelligent manufacturing: intelligent systems. This type of production line with only data kanbans can only be run by engineers, and there are many problems. There is even a dilemma where investment in this type of intelligent production line is "quick death", and no investment is "slow death". Mechanization ensures that products can be produced, automation enables products to be produced automatically (machine generation), digitization uses a large amount of data to facilitate review to ensure the realization of mechanization and automation, and intelligence ensures optimization

And unmanned operation (optimized machine generation). The stage of mechanization, automation, digitization, and intelligence that China and European and American countries are in. At present, many weaker domestic enterprises are still in the stage of mechanization, and the better ones are in the stage of mechanization, automation, and digitization. Although many problems can be seen through the data kanban, this is only a "make-up lesson" in the field of automation and digitalization. The quality of Chinese corporate data is their weakness. The machine generation stage in European and American countries has long been completed, and the production line mainly relies on automation rather than manual labor; and the digital needs of most European and American countries have basically been met. In terms of intelligence, many better European and American companies have better smart systems, and slightly worse companies are also trying to try smart systems. At present, China's smart manufacturing is mainly at the beginning and end of the manufacturing process, such as smart warehousing, delivery of incoming materials and finished products, etc., while the main process of smart manufacturing, such as the optimization of the production process, does not involve much. European and American countries are working hard to promote smart manufacturing. In intelligent manufacturing, a lot of work is incorporated into the intelligent system. Engineers or technicians continue to optimize this system mainly in the background; the optimized system is operated by on-site personnel to complete production; at the same time, managers pay attention to observe this system. In this way, everyone works together, and production continues to be optimized.

In such a situation, what kind of support policy should the Chinese government introduce?

After the implementation of the two standards, the government's industrial support funds have been more transferred to the field of intelligent manufacturing. The government's support for smart manufacturing should take into account the following points.

Focus on supporting intelligence (a distinction is made between automation, digitization, and intelligence. Many companies have already tasted the sweetness of automation and digitization. Even if the support in these areas is not strong enough, they will work for their own interests; but China’s intelligentization has just begun. In addition to computer systems and data collection, the requirements of intelligentization require a very in-depth engineering background. Therefore, the difficulty of intelligentization is much higher than that of digitalization. At the same time, due to the weak technological foundation of China, the In the field of chemical industry, the input-output ratio is relatively low in the short term, and government support is even more needed. Otherwise, the development of intelligentization in China will be slow.

vacuum coating machine,pvd coating machine,pvd vacuum machine,vacuum ion coating machine,multi-arc ion coating machine

It focuses on supporting the optimization of the production process in the core area (different from the peripheral areas such as smart warehousing). The core area of intelligent manufacturing has high technology content, great effect, and huge investment, so it is currently less; peripheral areas such as smart storage, AGV trolleys, etc., have gradually become the climate, and can already bring better profits to enterprises, even if they are not focused Support, enterprises will also work hard to advance this field.

Provide financial support based on the degree to which the data meets the high requirements of intelligent manufacturing. Encourage companies to focus on data quality: Digitalization should only receive financial support when it meets the data quality requirements of intelligent manufacturing; if the data is incomplete, or even fake data, it should not be supported.

Training funds focus on supporting trainers who have experience in smart manufacturing. Only those who have done smart manufacturing can teach students smart manufacturing; the government's smart manufacturing training fund should only support trainers with relevant training qualifications; those who do not understand smart manufacturing can also get a bunch of Baidu Introduce the importance of intelligent manufacturing and introduce other companies' "what" materials; this aspect must be eliminated, so that the government's training fund can really work.

Europe and the United States are in the stage of intelligence, while China's intelligence has just begun. The mechanization phase of most Chinese companies has basically been completed, and they are in the automation and digitization phase.

The machine generation is a typical automation. The digital data signage can make the status of mechanization and automation clear at a glance, so it has a strong application prospect. The mechanization and automation of European and American countries have already been completed, so machine generation is no longer a major problem. Except for a very small number of companies, the digitalization of European and American countries has also been completed.

In digitization, data quality is the key. Insufficient data quality of Chinese companies is a key obstacle. In addition to technology, management is the main factor hindering data quality, and most companies do not have adequate penalties for data loss. Intelligence is the most advanced manufacturing process that overrides mechanization, automation, and digitization. The requirements for data quality are far higher than digitization.

vacuum coating machine,pvd coating machine,pvd vacuum machine,vacuum ion coating machine,multi-arc ion coating machine

First of all, it is necessary to establish a model of a series of engineering problems or engineering parameters to be studied, which requires an extremely familiarity with engineering fields such as process, products, equipment and automation; secondly, it is necessary to collect high-quality online data and analyze the engineering model based on this data. Perform machine learning to make it fully tied to the production line. Then establish an intelligent system, combine the model of the production line and a large number of scene use cases into the intelligent system, and generate an optimized combination of production parameters that can be changed in real time, and handed over to the basic automation for execution.

It can be seen that intelligent manufacturing is the most optimized machine generation. At present, China only conducts a certain amount of smart manufacturing in the first and last parts of manufacturing, such as smart warehousing, but the manufacturing process is very rare, mainly by some platform developers. Companies with better technology in European and American countries have successfully applied intelligent manufacturing, while companies with average technology are still working hard.

China's main strategy for promoting intelligent manufacturing is to make up for digitization, automation and even mechanization, to encourage automation and digitization, and at the same time to encourage the development and implementation of intelligence. All Chinese enterprises must strictly attach importance to data quality in terms of technology and management, and have strict reward and punishment systems in terms of data quality. The ignorance of intelligent systems in the construction of intelligent production lines should be reduced. China currently has a lot of misunderstandings about what is truly intelligent manufacturing. Many digital manufacturing and even automated manufacturing, which are the necessary foundation for intelligent manufacturing, are regarded as intelligent manufacturing.

Founded in 2015,Zunhua Baorui Titanium Equipment Co.,Ltd. is a manufacturer specializing in pvd vacuum ion coating equipment. The company’s products mainly include large plate coating machine, large tube collating machine, tool coating machine and LOW-E glass production line. Mr.Wang baijiang ,general manager of the company ,has been engaged in vacuum coating industry for more than 30 years. He continuously improve production technology, improve product performance and devote himself to provide customers with better product experience and higher production efficiency.