In the depths of the Gobi in the Tarim Basin, geological engineers used to spend tens of days manually processing seismic wave data to determine the underground oil content. Now, relying on the green computing center and Kunlun large model built by enterprises, this cycle has been greatly compressed and the exploration efficiency has been significantly improved. This is not a science fiction clip, but a true epitome of China Petroleum's promotion of deep integration of artificial intelligence and energy. With the joint issuance of the Action Plan on Promoting the Dual Empowerment of Artificial Intelligence and Energy by the National Development and Reform Commission, the National Energy Administration, the Ministry of Industry and Information Technology, and the National Data Administration, and the holding of the National Artificial Intelligence and Energy Field Promotion Conference in Shenzhen, China's energy industry is entering a critical window period of intelligent transformation, with 51 high-value scenarios released. As the largest state-owned oil and gas production and supply enterprise in China, PetroChina firmly implements its deployment and regards artificial intelligence as a key variable driving high-quality development. From the foundation of computing power, business empowerment to organizational support, the three lines are advancing simultaneously, accelerating the construction of digital China Petroleum.

Building a green and low-carbon independent computing power base

China Petroleum is coordinating the layout of computing infrastructure and committed to building an integrated computing power system that is ubiquitous, networked, intelligent, agile, safe, reliable, green, and low-carbon. In the construction of green data centers, the comprehensive application of green electricity substitution, distributed new energy, efficient power supply and distribution, and water-saving refrigeration technologies will significantly improve energy utilization efficiency and create a benchmark for green computing power in the industry. In terms of computing power collaboration, we actively build an integrated microgrid of source grid load storage, accurately matching distributed green power resources such as wind power and photovoltaics with the inference computing power at the production end, achieving the generation and utilization of green power, improving the on-site consumption level of green power and the economic efficiency of computing power operation.

Deep cultivation of AI scenarios and application of Kunlun big model

Led by artificial intelligence and special actions, China Petroleum focuses on the intelligent upgrading of all elements of the oil and gas industry chain. In terms of high-value scenarios, we adhere to the five principles of highlighting the main business, reflecting value, knowledge intensive, rich language materials, and feasible technology. We focus on tackling scenarios such as intelligent control of gas storage peak regulation, intelligent calculation of exploration seismic waves, and integrated intelligent operation and maintenance of oil and gas production air space. We aim to create a pilot level intelligent factory demonstration and promote the large-scale replication of mature scenarios. In terms of dataset construction, we will deepen the implementation of data efficiency improvement actions, coordinate the construction of industry general and specialized datasets, build high-quality datasets with a scale of over 600TB, and establish a collaborative construction mechanism and market-oriented trading exploration. In terms of Kunlun Big Model, we will continue to strengthen our basic abilities such as intelligent question answering, logical reasoning, and multimodal understanding, break through advanced capabilities such as intelligent agent task planning, tool scheduling, and data and mechanism fusion computing, and overcome key technologies such as autonomous evolution and operator adaptation. In terms of security protection, we will build a security governance system that covers the entire lifecycle of data, models, and applications, strengthen the protection of critical infrastructure and the management of the entire lifecycle of data, and ensure that artificial intelligence applications are compliant, transparent, secure, and trustworthy.

Establish a sound organizational and talent innovation guarantee mechanism

China National Petroleum Corporation (CNPC) has established three major projects, namely artificial intelligence and special actions, to coordinate information technology strengthening, digital empowerment, and intelligent development. It has set up command centers for artificial intelligence big models, embodied intelligence, AI4S, etc., with the leader in charge of the Party group serving as the overall commander, and leveraging the professional support of various units' artificial intelligence work teams. At the same time, focusing on the development of embodied intelligence, high-quality datasets, and open typical scenarios, we will collaborate with universities and research institutions to build an innovation consortium for industry university research and application. Through the artificial intelligence community, we will promote the co construction and sharing of non confidential and non sensitive data and models. Adhere to the combination of cultivation and introduction, rely on the integration platform of industry and education to cultivate compound talents who understand both energy and artificial intelligence, implement the system of unveiling and leading, and make talents the largest increment of high-quality development empowered by artificial intelligence.

Looking towards the future, China Petroleum will take the implementation of the Action Plan on Promoting the Dual Empowerment of Artificial Intelligence and Energy as an opportunity, adhere to the original mission of safeguarding national energy security, continue to deepen the integration of artificial intelligence and special actions, and form a number of industry-leading innovative achievements, contributing petroleum strength to the construction of an energy powerhouse.Editor/Yang Meiling