Rewrite the Technological Landscape

On September 10, the state officially promulgated the "Three-Year Development Suggestions for Industrial Upgrading", suggesting that industrial upgrading needs to be more detailed, and it is necessary to focus on supporting some technologies that have achieved initial results.

On the same day, Dafeng Group invited hundreds of representatives of domestic manufacturing companies including automobiles and consumer electronics to participate in a conference on cloud intelligence + manufacturing.

The conference was hosted by Meng Qian himself.

"In the past two days, many old friends have called me to ask me what the purpose of holding such a meeting is, after all, the Gale Global Developers Conference has just ended less than 3 months ago.

In fact, it is because we have communicated with more than half of the top 500 global manufacturing companies in the past two months of the 4th Gale Global Developers Conference, and made a comprehensive summary and analysis of various details.

During this process, we noticed that the whole world now knows that the next key development direction of our Dafeng Group is cloud intelligence + manufacturing.

But now users’ understanding of this thing is still more in products, such as smart home appliances and smart cars, but everyone who came today knows that cloud-connected + manufacturing is not simply taking out a smart product and empowering traditional products. Smart systems and networking capabilities so easy.

The fundamental purpose of cloud-connected + manufacturing is to upgrade the industry, so today when the "Proposal" is released, we decided to invite everyone to come over and talk about this matter with the help of the market situation we have learned over the past two months. "

Meng Qian said that this opened the PPT, and everyone's attention was also focused, "We noticed some current situations in the process of contacting traditional manufacturing companies at home and abroad.

The foundation of traditional manufacturing originated from large-scale standardized production in the industrial age. The management model is based on pyramid, multi-level, and subdivision. This management model has poor flexibility and is difficult to adapt to rapidly changing manufacturing tasks and customer needs.

At the same time, there are too many subdivided fields in the manufacturing industry, and the industry standards for each subdivision are different. When Yunzhilian enters the manufacturing industry, there will be no standards.

For example, one of our most common problems is that enterprise workshops often have a large number of digital machine tools and industrial automation products of different brands, and design various industrial Ethernet and fieldbus standards. The software and hardware of manufacturers are even more difficult to be compatible. Traditional manufacturing This lack of relevant standards and complex production line status are hindering the development of cloud-connected + manufacturing.

Therefore, in order to adapt to cloud intelligence + manufacturing, almost all traditional manufacturing companies need to undergo a disruptive transformation.

However, this is a huge investment for enterprises, which will involve a large number of equipment, software and hardware updates and even transformations. The investment cycle is long and it is difficult to achieve results in the short term.

We can see an interesting statistic. Since the beginning of this year, more than 10,000 companies have come to discuss with us about cloud intelligence + manufacturing. Less than 200.

I don't know what you think when you see this data? "

Meng Qian deliberately paused to wait for everyone to respond, only to hear someone say that there is a long way to go, and then Meng Qian continued, "Yes, it is true that there is a long way to go, so when I first saw this When I reported it, I was so excited that I almost didn't sleep all night.

Because we are in contact with global manufacturing companies, what we collect is global manufacturing information, and what we draw is global manufacturing conclusions. That is to say, this hesitation is not only an exclusive problem of our Chinese manufacturing industry, but a global Common attitudes in manufacturing. "

When Meng Qian said this, the eyes of many people began to change, "From another perspective, isn't this our opportunity? When others are hesitating, isn't it the time when the opportunity comes.

Of course, the premise is that the technology is trustworthy, so today I will show three successful application directions from the practice itself. This is the relatively successful and even mature application of Yunzhilian in the manufacturing industry.

The first is smart detection.

In our communication with the international auto giants, our intelligent detection has become the focus of the whole industry. The manufacturing process of the automobile industry is extremely complicated, and the online detection task is extremely heavy.

But for a long time, everyone has always been mainly using manual detection. The results are obvious. The accuracy of manual recognition is very limited. It is not easy to have false detections, and the detection speed is also slow.

Coupled with the high mobility of testing workers and the difficulty in accumulating experience, major car companies have to invest a large amount of money in training every year.

However, the intelligent detection system launched by Dafeng Group has achieved very significant results in BYD and Geely factories. We record the production process through industrial cameras and hand over the video to artificial intelligence for machine detection.

At the beginning, our artificial intelligence needs to conduct double inspections with workers to achieve double insurance purposes. As artificial intelligence continues to accumulate inspection experience, deep learning begins to play a significant role.

So far, the artificial intelligence we use in BYD has replaced 50% of the inspection work of workers, and the detection rate of defective products is as high as 86%, and this data is constantly being optimized with the accumulation of experience. "

Meng Qian said that this is the beginning of the video to show the application of intelligent detection in BYD, giving everyone a more intuitive feeling.

"The second relatively mature technology application is intelligent maintenance. All factories are well aware of the importance of equipment maintenance, but in traditional factories, everyone basically adopts passive maintenance, and waits for equipment to go wrong before maintenance.

Now, the artificial intelligence-based intelligent maintenance we create can use machine learning to realize early warning of equipment maintenance. We also have a case here. During the process of our cooperation with Gree factory, the average number of overhauls of equipment has been reduced by 51%. System diagnosis and maintenance Response time is less than 1 hour.

It not only shortens the equipment maintenance cycle, but also improves the utilization rate of equipment. "

Next is naturally a video display, "Finally, let's talk about the third place where smart applications are effective, and that is the smart supply chain.

In the process of the globalization of Huaxia enterprises, we not only realized the importance of the vertical industrial chain, but also felt the importance of the supply chain.

The incident in the Neon Kingdom this time is believed to have brought a positive impact to many companies, and therefore many people are wondering why the Gale Group seems not to have been affected by this incident.

Today is also the first time I have responded to this question directly. In addition to our high self-sufficiency rate in the industrial chain, the key to our seemingly easy response to this incident this time is actually the smart supply chain we built internally. system.

For multinational companies like us, traditional supply chain management has shown very obvious defects in our globalization process, such as low efficiency, high circulation costs, inaccurate demand forecasting, insufficient supply response, and insufficient ability to deal with supply chain fluctuations. Manufacturers' inventory management costs are high and so on.

When we let machine learning enter the supply chain management, artificial intelligence can effectively establish a real-time supply chain matching relationship through the analysis of demand, planning and inventory. Through artificial intelligence, we have established multi-level inventory and planned production inventory dynamics Adjustments and even semi-automation of procurement and replenishment.

Let's take a look directly through the video. In this Neon Country event, our intelligent supply chain system immediately proposed to us the material procurement plan, the global factory production plan, the adjustment of the market supply plan for various parts of the world, and the future supply and demand forecast.

Through this system feedback, we immediately clarified the supply and sales targets of different countries and cities in the next two to three months, timely dispatched, and minimized the impact of the neon country incident on our company. "

The eyes of the representatives of these manufacturing companies that Meng Qian showed off with this set of intelligent supply chain system were shining.

"After seeing this, are you more interested in cloud intelligence + manufacturing?" Meng Qian asked with a smile looking at everyone's expressions.

Everyone nodded subconsciously.

"Then let's sort out the core technologies of cloud intelligence + manufacturing, which are semiconductor chips, core equipment components, core software and core algorithms.

Looking at it now, we are confident that we can compete with the United States in terms of core algorithms and software. We still lag behind the United States in terms of industrial semiconductors, and the biggest gap is still in core industrial equipment.

Do you agree with our judgment? "

Everyone nodded again, and Meng Qian also nodded, "Obviously, the comprehensive development of cloud intelligence + manufacturing is inseparable from these four cores, so after three successful application cases, let's discuss from these four cores. The state of the art."