Thread Content
On November 27, six departments including the Ministry of Industry and Information Technology announced the list of China’s first 15 flagship smart factories, with the petrochemical industry represented on this list. For petrochemical enterprises, how much room is there for digital transformation? What improvements can be made? What kind of effect will it achieve? At the sub-forum on \"Digital Transformation and Improvement of Petrochemical Enterprises\" held during the 2025 Petrochemical Industry Digital Transformation Conference in Dalian recently, some companies shared their experiences. Global optimization of smart factories: \"Enterprise digital transformation is inseparable from the development of smart factories, and digital twin factories represent an important component of digital engineering in the construction of such factories.\" ”Chu Xiangping, a senior expert at Sinopec Zhenhai Refining & Chemical Co., said, “With the overall goal of building a ‘collaborative’ digital plant, we have achieved simultaneous delivery of the physical plant and the digital plant.” Leveraging the engineering construction management information system, full coverage of project phases, complete integration of business processes, and comprehensive process management have been achieved. ”Sinopec Zhenhai Refining & Chemical is the only petrochemical company on the initial list of pilot smart factories. By continuously working on the development and application of 3D digitalization, Zhenhai Refining & Chemical has created a digital twin application system that is now being effectively utilized in tasks such as model review, design boundary verification, pipeline construction management, quantity settlement, asset delivery, and pipeline process handling. During the operation of smart factories, how to achieve overall optimization in order to maximize efficiency across the entire chain – from raw materials/data, through processes and flow control, to process equipment, planning and scheduling, and finally products – while minimizing costs, is a common challenge faced by petrochemical companies during their digital transformation. To this end, Xiao Xin, a researcher at the Institute of Process Engineering, Chinese Academy of Sciences, presented solutions in his report. “Global optimization is the intelligent underlying technology that spans molecules, raw materials, equipment, processes, and factories; it helps to foster new forms of productivity in the petrochemical industry, enables coordinated scheduling of large-scale production, maximizes the overall efficiency of refining operations, and simultaneously meets **major strategic needs. For example, the new generation of intelligent global optimization solvers can be successfully used for the global optimization of the wet phosphoric acid process, enabling precise multi-scale simulation and optimization across the entire process. ”Xiao Xin provided a detailed overview of the research achievements in next-generation intelligent global optimization, and discussed breakthroughs such as the definition of feasible solution sets, convex neural network estimators, and hyperplane branching. Building a modern and intelligent enterprise management system is crucial in the process of constructing and operating smart factories in this industry. As an important strategic base for energy and chemicals in China’s western region, CNPC Dushanzi Petrochemical Company has accumulated rich practical experience in the field of intelligent manufacturing, and was awarded the title of “Model Enterprise in Intelligent Manufacturing” in July this year. “Enterprises should establish a comprehensive management system, including a quality management system, a safety management system, and an equipment management system, among others. The various systems are interrelated and interdependent, which is crucial for business management. ”He Wei, a senior supervisor in the Mechanical Equipment Department of Dushanzi Petrochemical Company, said that the company adopts an integration model of \"data + platform + modules\" to create accurate profiles of its various units. Through a framework consisting of \"1 dashboard + 7 models + 3 data modules + 7 management processes\", it manages and standardizes the various elements of its system. This approach has completely transformed the company’s traditional experience-based management, which was based on the number of employees, and has enabled the transition of equipment management towards an information-driven integrity system centered on \"reliability + cost-effectiveness\". Artificial intelligence reshapes value. AI possesses typical characteristics of general-purpose technologies; how can it be integrated deeply into the petrochemical industry? “As the first industry-scale large model in China’s energy and chemical sector to undergo **registration, Kunlun Large Model has accurately identified the intersection point between artificial intelligence and industrial development. ”Hou Hongyu, an engineer at the Basic Research Center of CNPC’s Digital Intelligence Institute, said, “The Kunlun Large Model has identified five key objectives: industry-wide large models, application scenarios, datasets, AI platforms, and computing centers. By adhering to a business-driven approach supported by technological innovation, 470 intelligent applications have been designed, resulting in a comprehensive framework of ‘ten domains, hundreds of scenarios, and thousands of applications’.” ”At present, CNPC has developed 62 large-scale models that cover various business areas such as exploration and development, oil and gas production, refining and petrochemicals, oil and gas sales, and equipment manufacturing. These efforts have begun to yield results in driving innovation within the industry, thereby significantly improving operational efficiency. At the same time, as the hub for managing the development of artificial intelligence in China’s petroleum industry, the Kunlun·AI platform has been widely deployed, enabling the unified management of a range of tools used for training and deploying large-scale commercial models. Regarding the value of large models in the petrochemical industry, Li Lin, Chief Architect for the energy sector at Beijing Zhongshu Ruizhi Technology Co., Ltd., shared insights on the technological evolution of AI Agents and their practical application in industry. Li Lin pointed out that by integrating digital technologies across the entire value chain of the oil and gas industry – including exploration and development, production and operation, as well as storage, transportation, and sales – AI Agents enable low-carbon operations and innovative service models, thereby unlocking new potential for cost reduction, efficiency improvement, and value creation in this industry. “We focus on computing power, data, and algorithms to build domestic computing power clusters, and use AI to drive corporate management upgrades as well as improvements in business efficiency. We have already integrated mainstream domestic large models into our capacity-building efforts, with the aim of creating a secure, controllable, and uniformly managed AI platform to support enterprises’ digital transformation. ”Huang Hongtao, head of the Digital Management Department at Zhonghua Chemical Digital Technology Company, said, “We have developed mature products that lead industry innovation in areas such as AI-based material layout and planning, AI+CAD-assisted design, and AI-driven material modification. The potential for applying artificial intelligence is enormous, and our company will continue to explore applications in the engineering construction sector through an AI+ strategy.” ”Given the diverse scenarios and requirements across various sub-sectors of the petrochemical industry, and the fact that existing control technologies are insufficient to meet the needs for autonomous optimization under all operating conditions, Hou Weifeng, director and deputy general manager of Zhejiang Zhongzhida Technology Co., Ltd., proposed a new generation of intelligent factory solutions based on AI and the industrial internet. These solutions can be applied in various scenarios such as zero-manual operation, simulation and optimization of production processes, optimization of production planning and scheduling, energy balance and energy-saving scheduling, online equipment diagnosis, and AI-powered image recognition.