Zhu Weilie of Huaneng: Industrial Internet should shift from "platform-oriented" to "application-oriented"

　　In recent years, industrial informatization has been highly valued by the country. Under the guidance of policies and the joint efforts of relevant parties, the development of my country's industrial Internet has made a good start. Recently, Zhu Weilie, chairman of the Industrial Internet and Big Data Application of the China Computer Users Association and chief information officer of Huaneng Group, said in an exclusive interview with Xinhuanet that the development of my country's industrial Internet should shift from "platform-oriented" to "application-oriented".

　　The following is the interview transcript:

　　Xinhuanet: What is the current overall development of my country's industrial Internet? What are the main problems facing the development of my country's industrial Internet?

　　Zhu Weilie: I think that although the industrial Internet has made considerable progress, it is still in its infancy overall. Now the enthusiasm of all parts of the country is relatively high. Relevant departments have issued guiding opinions and various policies on the industrial Internet, and various provinces, cities and autonomous regions have also issued many documents to provide support. At the same time, many information technology companies and Internet companies are also highly involved, which is a good situation we have seen.

　　The first problem facing the development of the industrial Internet is the lack of participation of industrial enterprises. Currently, the main companies involved in the construction of the industrial Internet are information technology companies and Internet companies. However, these companies have never done industrial Internet in the past, and they do not have a deep understanding of industrial business. The industrial Internet business involves the production process of industrial enterprises. Without an understanding of industry, it is difficult to develop good intelligent applications. Therefore, in the future construction and development of the industrial Internet, the active participation of industrial enterprises is crucial.

　　The second problem with the development of the industrial Internet is that it is too "platform-oriented and application-light". There are many companies developing industrial Internet platforms in the market, but few companies are truly developing intelligent applications for the industrial Internet. The reason for this situation is that information technology companies and the industrial Internet generally do not understand industry and can only focus on platform development. Some information technology companies actually use the "platform concept" to cover up their lack of capabilities in the industrial Internet. Another reason is that due to the influence of the past TO C business, in the TO C business, the Internet platform has been sought after, and people think that the platform is also the core of the industrial Internet business, so they compete to develop industrial Internet platforms. In fact, the development of the industrial Internet is just the opposite. Innovative intelligent applications are the core. Due to the current situation where few companies that develop the industrial Internet use it, and even fewer good applications, the participation of the entire industrial enterprise is not high, which creates a vicious cycle.

　　Therefore, in addition to encouraging industrial enterprises to participate as much as possible, we also hope that information technology companies will strive to achieve transformation and develop better application products based on industrial realities.

　　Xinhuanet: What is the significance of developing the industrial Internet in promoting the digital transformation of the real economy?

　　Zhu Weilie: The idea of ​​achieving high-quality development of traditional industries through the integration of industrialization and information technology has been proposed for many years, but this road has always been difficult. On the one hand, in the past few years, our informatization has basically remained in the field of enterprise management informatization, such as office automation, ERP, etc. These informatization contents are mainly to improve the standardization of enterprise management. Although they also involve the production management of enterprises, the integration of industrial production processes is not deep enough. The industrial Internet uses production big data and directly improves the production efficiency of enterprises, such as optimizing energy saving and consumption reduction, providing equipment status maintenance, etc. It can be said that the industrial Internet has directly promoted the high-quality development of the real industry.

　　On the other hand, we have always adhered to the concept that the core content of the industrial Internet is to integrate industrial big data and artificial intelligence into traditional industries. Therefore, the industrial Internet and intelligent industry may bring a new round of revolution to the production and operation of industrial enterprises. Foreign companies such as GE (General Electric Company of the United States) and Siemens have tried to invest billions of dollars to occupy this technological high ground. At present, most domestic companies have no personal feelings about this. For example, we now need to buy domestic coal and imported coal for thermal power generation. However, there is not only a high price difference between international coal prices and domestic coal, but also different volatility patterns. Therefore, how to purchase imported coal and domestic coal has a great impact on the operation of enterprises. Relying on machine learning and artificial intelligence to innovate in this regard has a very large impact. Traditional industrial enterprises often have no time to take care of new technologies because they are busy with production work all year round, but as scientific and technological workers, we must connect modern artificial intelligence technology with traditional industries and promote the transformation and upgrading of traditional industries through industrial Internet and industrial intelligence.

　　Xinhuanet: Your team uses digital twins to implement condition-based maintenance of equipment. How is this possible? What is its significance?

　　Zhu Weilie: In the past, my country's industrial field has always adopted regular maintenance and periodic overhaul. Condition-based maintenance has been proposed in the industry for decades, but it has not been truly solved in practice.

　　Now we are exploring condition-based maintenance by relying on digital twin technology. The concept of digital twin is to calculate the digital model of the equipment through machine learning and artificial intelligence technology. This model has two characteristics: one is personalization, each device has a digital twin, and the digital twin models of industrial equipment of the same model are different; the other is that the model changes synchronously with the wear and tear of the equipment. In the future, we can use the digital twin of each device to fully grasp its health status, and use it to set the maintenance time of the equipment to achieve personalized condition-based maintenance.

　　In the industrial field, aviation, petrochemical, steel, metallurgy and other industries can all use this technology to achieve equipment status maintenance.

　　Xinhuanet: Can you introduce one or two successful cases?

　　Zhu Weilie: Let me first introduce the intelligent applications of our thermal power plant.

　　The first is the desulfurization optimization control system. We developed this system in a power plant. Through model calculation analysis and autonomous learning, it finally outputs the limestone slurry supply valve opening instruction, optimizes the slurry supply of the absorption tower, and forms a closed-loop regulation control with the sulfur dioxide concentration at the discharge port. This application is expected to save about 750,000 yuan in material consumption per unit of desulfurization equipment per year, and about 100,000 yuan in plant electricity per unit of desulfurization equipment per year. If this intelligent application can be promoted nationwide, it can generate about 3 billion yuan in benefits each year.

　　The second is boiler tube safety monitoring. Four-tube leakage in thermal power plants is the primary threat to boiler safety operation. We use AI intelligent algorithms to establish a data model of the correlation between boiler tube temperature and boiler operating conditions, as well as a data model of the temperature and position between boiler tubes, to achieve real-time analysis and historical change analysis of the temperature of the boiler tubes, give the cause of overheating and provide early warning. During production operation, the operation mode can be adjusted in time to improve or delay the development trend of faults. This technology is really helpful in reducing non-stop operations due to four-tube leakage in boilers.

　　We have been exploring the industrial Internet for seven or eight years. Although these years of exploration have been bumpy, we have indeed seen that this technology has great potential to improve traditional industries. We hope that through our continuous exploration, we can truly embark on a path of high-quality development of traditional industries.