Industrial Internet "Heroes And Heroes" Jia Yongli Explains HUAWEI Yun Fu Way

Song Wenhui diagram

In the fourth wave of industrial revolution, the industrial Internet as the key technology is of great significance for human-machine interaction and industrial digitalization.

In the face of this industry highland, dividends on domestic policies are frequent. Recently, the Ministry of industry and information technology issued the notice on Issuing the "5G+ industrial Internet" 512 project promotion plan. It is clear that by 2022, it will break through a number of 5G key technologies that meet the specific needs of the industrial Internet, and build a batch of "5G+ industrial Internet" Intranet to build benchmarking and model projects to form at least 20 typical industrial application scenarios.

The so-called industrial Internet can be understood as the ability to transmit cloud storage, data analysis and intelligent decision-making to various industries on the basis of industrial digitalization. In the meantime, 5G, AI, cloud computing and other technologies are needed to reduce the cost and efficiency of the manufacturing process.

Therefore, manufacturing giants are competing for the layout of the industrial Internet platform. From domestic HUAWEI, Foxconn to overseas SIEMENS and GE, they are exploring their respective paths. At present, the domestic industrial Internet market is still in the embryonic stage of development, and enterprises are scrambling for innovation.

Jia Yongli, President of HUAWEI cloud artificial intelligence, told an interview with reporters in twenty-first Century economic report: "many domestic enterprises, especially large enterprises, are very open to AI and 5G technology. The management of the enterprise is very concerned and willing to try, but at the same time, it is relatively prudent. Because industrial development has gone through many years and has gone through the process of information and automation, there are many industrial platforms and software on the actual production line, and the changes brought by industrial Internet and AI are not achieved overnight. And now the AI is also in the initial stage of exploration, so at the moment I think it is still tentatively used, is still in its infancy.

Industrial Internet Age

As mentioned above, the current pattern of industrial Internet is uncertain, and it is still the age of heroes.

Globally, participants on the card table are mainly industry giants, who can be divided into three categories. One is the international manufacturing giant, such as HUAWEI, SIEMENS, Foxconn and other large world-class manufacturing enterprises. They have accumulated a lot in the industrial process, and have formed their own system.

Nowadays, these giants want to open the whole system to enable large and even small and medium-sized enterprises in all countries to access their standards. At the beginning of November, HUAWEI industrial Internet platform FusionPlant was formally selected as a cross industry and cross domain industrial Internet platform. This also meant that FusionPlant became the benchmark for the construction and promotion of industrial Internet platforms in various industries.

The ambition of these giants is that they hope to export the industrial overall digital optimization capability, and at the same time, be able to get the scale effect of the whole system in the industrial field through lower marginal cost.

The second type is the industrial digital system that is applied only to the enterprise itself, that is, the digitization and meticulous management of the enterprise itself. The third category is a company that specializes in the third party industrial Internet solutions. These three types of enterprises or based on the original manufacturing foundation, or based on the accumulated software technology, customize industrial systems to vertical industries.

The commercial use of 5G brings greater opportunities to the industrial Internet. Industrial Alliance CEO Zheng Hong Meng, an example of the twenty-first Century economic report, said: "the future components are getting smaller and smaller (entering the nanometer level), but the industrial images with the resolution of 2K or even below 2K can hardly meet the demands of industrial development. The 8K image is 16 times the 2K's, which can better solve this problem. In this way, the complex and huge data flow needs to be transmitted to the cloud by the 5G network to achieve the goal of producing 100% yield. 5G can also reduce costs for enterprises. For manufacturing enterprises, reducing operating costs is equivalent to making money.

At the same time, the promotion of the industrial Internet platform is also facing the industry challenge.

Jia Yongli told reporters that the current industrial Internet has three main difficulties. The first is that some management expectations are too high, the rhythm is unreasonable, and there are too many things to do. A short time to do a particularly large industrial Internet system will face many challenges, not only in terms of technology, but also in organization, culture and process.

Second, the direction is not clear. "To do the industrial Internet, first of all, we need to be clear about what problems I care most, and what is the core issue of my industry." Instead of others doing quality inspection, you must also do quality inspection, because quality inspection is the easiest to see and the easiest to do, so many enterprises are doing it now. But is this the core of your business? So the selection of scene is a difficult point. Jia Yongli said.

Third is the problem of data output and collection. He explains: "the data of the enterprise is in the machine, in the equipment, in the existing system. In this sense, enterprises have a lot of data, but at the same time they are also faced with the lack of data. Because with the emergence of AI, in addition to the existing mechanism models, we also need to build some non mechanism models through data driving to match the mechanism models. However, data sets built by enterprises for existing mechanism models are not necessarily suitable for AI modeling. So the data that AI needs is not collected by the enterprise, or many core equipment may not be open to the public. This is also a matter for enterprises to focus on in the coming period.

HUAWEI cloud's "black land"

The foundation of a country's development of the industrial Internet is that its industrial level has reached the general digital level. The main bases include sensor control, data collection, extensive application of industrial standards, and even the growth of senior technicians. These foundations can not be achieved overnight. All this is exactly what China is gradually getting, but it has a long way to go.

Among them, HUAWEI, as the leading enterprise of ICT, is an important participant in promoting the industrial Internet. Its industrial Internet platform FusionPlant is based on HUAWEI cloud external energy.

"The difference between HUAWEI and other enterprises is that, first, HUAWEI's industrial Internet platform is the positioning of the platform. We are the enabling layer, helping other industry partners on my platform to do industrial APP in the era of artificial intelligence, 5G and IoT." Jia Yongli introduced to reporters: "secondly, HUAWEI has accumulated IT and CT technology over the years, we will integrate. In connection management part, we can integrate IoT, 5G and edge computing to enable enterprises to access all kinds of applications and sensors. On top of that, we have a complete tool chain based on cloud development, and we also have data, AI, processing, intelligent and open services.

While expanding the industry, HUAWEI's approach is to find benchmarking enterprises, or core integrators and equipment manufacturers. They have a deep understanding of the industry, while HUAWEI provides cloud computing, 5G and AI technology. After the cooperation forms the standard solution, then promotes to the profession.

At present, HUAWEI has gone deep into various industrial segments, and has landed cases in many fields, such as petroleum, chemical fiber, petrochemical, steel, coal coking, electronic equipment, automobile manufacturing and automatic driving.

For example, in the production line of HUAWEI Songshan Lake factory, each worker takes a few minutes to do the test every time, and it may misinspect and miss. The visual inspection model of HUAWEI cloud artificial intelligence service improves the efficiency of defect detection in manufacturing process, increases the yield of the production line to 99.55%, and reduces the AOI test staff by 48%.

For example, the synthetic fiber industry system integrator Sanlian Hong Pu, with the help of HUAWEI cloud industrial intelligence, helps fiber manufacturers to classify their products, accurately predict the quality of products, increase the demand matching rate by 28.5%, and test the fiber quality from 100 to 1 thousand meters per line of artificial silk.

HUAWEI cloud said that FusionPlant will continue to make "black land" in the industrial Internet field. "We call it" black land ", that is, we hope that these integrators will be on our land, they will" own "a variety of rich products, each industry has a very deep professional knowledge, so our path must be to find more partners. Jia Yongli said.