The hottest Siemens Yaojun in the third digital Ch

Siemens Yao Jun's keynote speech at the intelligent manufacturing sub forum of the third digital China Construction Summit

on October 12, Yao Jun, vice president of Siemens (China) Co., Ltd. digital industry group and general manager of process automation department, attended the intelligent manufacturing sub forum of the third digital China Construction Summit and delivered a keynote speech on the theme of industrial artificial intelligence + big data analysis: Unlocking the future of factory intelligent operation and maintenance. In his speech, Yao Jun focused on the equipment predictive maintenance system SIEPA developed by the Siemens China team

Siemens Yaojun's keynote speech at the intelligent manufacturing sub forum of the third digital China Construction Summit

with the support of artificial intelligence technology, SIEPA makes full use of the historical data of the factory, and through the equipment operation status prediction and early warning module and intelligent troubleshooting and diagnosis module, it can not only predict the failure risk in the early warning operation in time, but also help enterprises diagnose the cause of the failure efficiently and guide them to repair and maintain, So as to effectively control risks, reduce costs and increase efficiency. In China, SIEPA has been successfully applied in many customer factories, including Sinopec Qingdao refining and Chemical Co., Ltd. In the intelligent plant of Qingdao refining and chemical company, SIEPA has established a closed-loop mechanism from intelligent early warning to advanced diagnosis for customers to ensure the reliability and safety of production

Siemens Yaojun's keynote speech at the intelligent manufacturing sub forum of the third digital China Construction Summit

speech transcript:

today I hope to share with you how to apply artificial intelligence and big data to the industrial field to help industry implement the intelligent operation and maintenance of factories. From simple automation to digitalization, we are more about realizing the integration of the two, creating an end-to-end value chain, forming a closed-loop information flow from purchase, order, production scheduling, warehousing and logistics to users, and integrating all data from design, engineering, production, operation and maintenance to services into one platform. This Pu course material is developed on the basis of Pu course material. The new generation of course material is what we are doing today, which uses a lot of intelligent technology, but this is not the goal of the real intelligent factory we pursue. The intelligent operation and maintenance we hope to achieve is divided into three levels: one is to strengthen perception, from cognition to sensing to cognition; First, our production mode, business mode and development concept are changing from precision to best; One is to sharpen operations, from professional to collaborative

digital factory includes three digital twins: Digital twins of products, digital twins of production and digital twins of performance. Based on the digital twin technology, we can integrate the applications of enhanced perception, optimized control and sharpened operation in intelligent operation and maintenance into many application scenarios. In the process of product design, realize product quality early warning, risk prediction, parameter optimization and Simulation in product production, and predictive maintenance in performance, such as abnormal early warning and intelligent diagnosis

industrial artificial intelligence is the core tool for the implementation of factory intelligent operation and maintenance. Today, the artificial intelligence we encounter in daily life can also be used in intelligent operation and maintenance. For example, supervised learning is a good example. With the rapid development of the express industry. To complete the abnormal state warning in an industrial scenario, you first need to identify what is abnormal, and the system will give an early warning after the abnormality occurs. If anomalies occur frequently, the system will predict the risk trend, etc. Reinforcement learning is the same. Through continuous self-learning, we can apply optimization control to the factory, such as realizing parameter optimization. Production includes process, and there are many correlation analysis can realize the optimization of parameters, so as to improve quality and efficiency. Then the highest level is the knowledge system and knowledge map. Just like when we go to see a doctor today, we used to be curious. Today, we need to do a lot of examinations, and we need the judgment of doctors of different majors to help doctors make the final diagnosis. In the industrial scene, it is necessary to form a knowledge map. If there is a problem with a device, experts need to see it and make a diagnosis to solve the problem. Today, we can solve this problem with knowledge map, that is to say, after accumulating a lot of information and knowledge, if there is a problem with equipment, we can compare it in the knowledge map. Has it happened in the past? How did it happen? What caused it? How was it solved later? In other words, we can no longer rely on an old expert, but a knowledge base to help us quickly find the key to the problem and solve the problem

industrial big data analysis is also an important foundation for intelligent operation. Today, most of the use of data is still in the application of data description. The dazzling large screen display we see in the factory is still some real-time data or filtered and summarized data. Real intelligence is continuous learning and optimization. There are two main stages for a factory to learn and optimize independently. One is data analysis. Through in-depth analysis and prediction, a lot of information can be provided to plant operation and maintenance personnel. The second is the full life cycle data closed loop. If there is a model, the model will continuously feedback all the data obtained in the future to optimize the model and form a closed loop. Feedback is very important to enable the system to achieve self-learning and self optimization

today, Siemens China team has developed an important predictive operation and maintenance platform SIEPA, which currently has two major modules, one is state prediction and early warning, and the other is intelligent troubleshooting and diagnosis. The first early warning is to predict the possible problems of key equipment in advance. Lei Wen, director of the environmental protection department of the Department of energy conservation and comprehensive utilization of the Ministry of industry and information technology, also said that the diagnostic module will explain where the problems may arise and how to solve and avoid such problems, which can help customers reduce unplanned parking and improve efficiency

we start machine learning from industrial big data, find the correlation between various parameters and establish models, then establish risk prediction and early warning, provide intelligent analysis and diagnosis, and then iterate and optimize the model through new data feedback in the whole life cycle to further help the machine learn. The continuous improvement of life cycle will greatly improve the availability and operation efficiency of the plant, including correct decision-making, cost reduction, etc

as a successful attempt of AI product design in industrial scenarios, SIEPA won the 2020 German Red Dot Design Award and was selected into the top 30 list of outstanding AI leaders (SAIL) at the 2020 world AI conference. SIEPA has been successfully applied in many customer factories, including Sinopec Qingdao refining and Chemical Co., Ltd., and has established a closed-loop mechanism from intelligent early warning to advanced diagnosis for customers to ensure the reliability and safety of production. At present, Siemens has carried out relevant cooperation and applications with many enterprises based on SIEPA in the global market, and is constantly moving towards the goal of digital and intelligent manufacturing

keynote speech of Siemens Yaojun at the intelligent manufacturing sub forum of the third digital China Construction Summit