How should we understand industrial big data, how can industrial big data enable intelligent manufacturing, and how to move toward the future industrial Internet on the basis of intelligent manufacturing?

The path of transformation and upgrading of industry

5X00062G01 The National Engineering Laboratory of Big Data System Software summarizes the path of industrial transformation and upgrading as four quadrants of “addition, subtraction, multiplication and division”.

The so-called addition and subtraction is intelligent manufacturing. Intelligent manufacturing is more concerned about the internal things of the enterprise, intelligent manufacturing in a narrow sense focuses on manufacturing, that is, the production link, and intelligent manufacturing in a broad sense includes the whole life cycle of the enterprise, from research and development design to production and manufacturing to operation and maintenance services.

5X00062G01 Intelligent manufacturing is nothing more than adding some things to the existing process and reducing some things, which can basically be summed up in eight words: improve quality, increase efficiency, reduce cost, and control risk. Today, smart manufacturing does things like add and subtract.

However, in this era, it is not enough to add and subtract, for example, private equity institutions invest in a company, the company does a little addition every year, investors may not be satisfied, but want the company to achieve exponential growth.

5X00062G01 How to achieve it? The industrial Internet could be the way to multiply and divide. Multiplication is the platform effect. Taobao, for example, which hosts countless stores that set up shop on its platform to make money, is a case in point. But in the industrial sector, can you build an industrial Internet platform?

Take the garment industry as an example. The traditional first generation of clothing enterprises have their own design, factories, stores, that is, a complete industrial chain. The second generation of clothing enterprises, to abandon the factory to choose full OEM production, to do marketing, with stores as assets. Clothing enterprises in the Internet era, neither factories nor stores, the cost is almost zero, all stores rely on Taobao, only responsible for rapid design, control of the supply chain, the final “total plate” although not necessarily as large as traditional enterprises, but high profit margins. Therefore, division is to focus on their core competitiveness.

Asset-light and high-profit operation is the way of innovation and entrepreneurship for Chinese smes in the future. Building the platform ecology of the industrial Internet does not mean that only this platform can make money, but everyone on the platform can make money.

Three levels: Classification of industrial big data industries

5X00062G01 The National Engineering Laboratory of Big Data System Software has contacted and done a lot of industrial big data applications, and it is divided into three levels.

The first level is the unit level, that is, for industrial equipment, not only limited to the remote operation and maintenance of equipment, but also includes early warning of equipment failure, fault analysis, and optimized operation of equipment, asset management, and so on.

5X00062G01 First of all, we need to carry out accurate digital measurement of the operating state of the equipment, which is actually the continuous spatial discretization of industrial big data. This continuous space is very complex, and the physical quantity, accuracy, and number of sensors that can be measured are limited, so full-space sampling cannot be realized. However, with the improvement of the digital level, the advancement of information technology and the iteration of intelligent applications, the future measurement process will also be upgraded.

5X00062G01 The second level is the factory level. This level does not focus on individual equipment, but focuses on the operational efficiency, product quality and safety, and environmental protection of the entire plant. Industry focuses on factors including people, materials, processes, equipment, and the environment, which can work together in complex dynamic systems.

If the whole of China is regarded as a big factory, how can we improve our efficiency in the industrial chain? Today, we work in big data and do “intelligent +”, which is the purpose.

The first thing to answer is where is the data, which is anywhere. In the past, the industrial data management is relatively rough, and the traditional management information is relatively good in the field of information management, and now a lot of industrial data is only used to do monitoring and do data playback when the fault occurs. How these data are used for the integration of the two (information and automated data fusion) has yet to be verified.

Level three is how to get other relevant data? For example, to automate the construction of excavators, it is necessary to understand GIS data and environmental data, but these are not the data owned by traditional manufacturing enterprises. This shows that the connotation of today’s industrial big data is much larger than the traditional data connotation. Automation and cross-border overall data constitute the system of industrial big data.