财经频道更多独家策划、专家专栏，免费查阅>> 责任编辑：山上 据"微讯江苏"公众号，2月1日，重大基础设施工程材料全国重点实验室理事会2026年第一次会议及2025 年度学术委员会在南京召开。现场，东南大学全球首发"砼真砼知"混凝土材料科学大模型。该大模型已 应用在南京北站建设项目中。该模型由重大基础设施工程材料全国重点实验室联合东南大学大数据中 心、阿里云计算有限公司共同打造，是国内首个混凝土材料科学大模型。 ...

Group 1 - The core viewpoint of the article highlights the launch of the "Concrete True Knowledge" model, a significant advancement in concrete material science, which was unveiled at the first meeting of the National Key Laboratory of Major Infrastructure Engineering Materials in 2026 [1][3] - This model, developed by the National Key Laboratory of Major Infrastructure Engineering Materials in collaboration with Southeast University and Alibaba Cloud, is the first of its kind in China, integrating artificial intelligence with concrete material science [3] - The model aims to revolutionize the traditional paradigms of engineering material research and application, positioning China at the forefront of intelligent engineering materials, and providing a "Chinese solution" for the digital and low-carbon transformation of global infrastructure [3] Group 2 - The "Concrete True Knowledge" model offers expert-level, traceable intelligent Q&A capabilities, addressing complex issues in concrete material research, design, and engineering applications [3] - It incorporates four knowledge processing tools: intelligent data extraction, multi-dimensional data comparison, literature review generation, and knowledge graph construction, significantly reducing the time required for literature research and data analysis from days or months to minutes [3] - The model features an open intelligent platform that includes three core intelligent bodies for performance prediction, mix design, and crack risk assessment, creating an end-to-end closed loop for complex problems reliant on experience [3][5] Group 3 - The model has already been applied in the construction of Nanjing North Station, assisting in the design of low-carbon concrete materials and predicting the hydration and strength development of concrete based on real-time environmental conditions [5] - It represents a shift towards a data-driven "fourth paradigm" in concrete material research, providing benchmark solutions to critical challenges in data, algorithms, and models within the engineering materials sector [5] - The implementation of this model is expected to accelerate the low-carbon transformation and high-quality development of the building materials industry [5]

Core Insights - The "Concrete True Knowledge" model, developed by the National Key Laboratory of Major Infrastructure Engineering Materials in collaboration with Southeast University and Alibaba Cloud, is the first concrete materials science model in China, marking a significant advancement in the field of intelligent engineering materials [3][5] - This model integrates artificial intelligence with concrete materials science, enhancing the accuracy of performance predictions and providing a digital, low-carbon transformation solution for global infrastructure construction [3][5] Group 1 - The model can unify various influencing factors into a computational framework, significantly improving the accuracy of concrete performance predictions [3][5] - It features a "deep specialization" approach, allowing for expert-level, traceable intelligent Q&A, and integrates four knowledge processing tools to streamline literature research and data analysis from days or months to minutes, increasing efficiency by over 90% in design phases [5][6] - The model includes an open intelligent agent platform that offers three core intelligent agents for performance prediction, mix design, and crack risk assessment, filling a technological gap in intelligent concrete design [5][6] Group 2 - The model benefits a wide range of users, from the general public concerned with housing construction to professionals in engineering management and researchers in advanced exploration [6] - It has been applied in major projects, such as the Nanjing High-Speed Railway North Station, where it adjusts concrete performance predictions based on real-time environmental conditions [6] - Future plans include continuous improvement of the model, development of specialized intelligent agents, and integration with production and construction equipment for real-time sensing and automatic adjustments throughout the entire process [6][7]