On July 23, the Municipal Bureau of Industry and Information Technology announced that the Ministry of Industry and Information Technology recently released the "2025 List of Typical AI Application Cases." Spanning five major directions—technology infrastructure, industry empowerment, product applications, support assurance, and thematic topics—285 cases were selected for 2025. Among them, seven projects from Nanjing made the list.
These include projects like the "Innovative Application of a Domestic Full-Stack Power-Specific Large Model Toolchain" by Guodian Nanjing Automation Co., Ltd., the "Practice of Intelligent Unmanned Crane Solutions Based on AI Dynamic Optimization Algorithms" by Nanjing Keyuan Wisdom Technology Group Co., Ltd., and "Photon Intelligent Engine OptoChat Assisting Photonic Chip R&D and Design" by Nanjing Nanzhi Advanced Optoelectronic Integration Technology Research Institute Co., Ltd. What sets these cases apart to earn national recognition? Our reporter conducted an in-depth investigation.
High Technological Content
Photonic chip R&D and manufacturing is a highly specialized and interdisciplinary engineering task. In real-world research scenarios, engineers and scientists often face vast amounts of literature, patents, process data, and application cases. Finding credible information quickly, accurately understanding parameter relationships, and converting scattered experience into reusable R&D pathways have long been challenges. To address these pain points in the photonic chip industry, Nanzhi Optoelectronics independently developed the photonics-specific large model OptoChat AI. This model deeply integrates photonic chip-related patents, literature, industry data, and trusted process knowledge to build a specialized knowledge system. It is not a general-purpose Q&A tool for all problems, but a professional intelligent assistant for the entire photonic chip industry chain, as introduced by a member of the R&D team. Currently, OptoChat AI is continuously iterating in areas like photonic chip layout design, process verification, yield management, and R&D support, evolving from a "knowledge Q&A tool" toward an "intelligent assistant for photonic chip R&D and production," gradually forming a closed loop of "professional data—model capability—process scenarios—industrial application." Being selected as a national benchmark case is not an end but a new starting point. A company representative stated that Nanzhi Optoelectronics will continue to advance OptoChat AI's iteration and upgrade, enhance its practical value in photonic chip R&D design, process verification, production, and industrial services, and promote deeper integration of AI with the photonic chip industry, helping to drive high-quality development of China's optoelectronic integrated circuit industry.
Nanjing Fiberglass Research and Design Institute is also demonstrating sustained innovation in new materials, making the list with its "AI-Empowered High-Performance Glass Fiber R&D" project. The institute has deeply integrated the concept of material genome engineering with glass fiber R&D, building a full-chain technical system of "molecular simulation—AI-driven—efficient verification." The team has established high-precision machine learning prediction models for key performance indicators such as glass fiber modulus, density, viscosity, and dielectric properties. The multi-objective optimization design technology integrating composition, process, and performance has reached an internationally leading level. This technology has already been industrialized in glass fiber enterprises, reducing R&D costs and shortening development cycles by over 50%, making it the first example in China's domestic glass fiber industry to achieve full-chain material genome engineering application and industrialization.
Empowering Traditional Industries
In traditional industrial settings, heavy-load handling operations by workers face significant safety risks, low production efficiency, and chaotic inventory data. Keyuan Wisdom has developed a full-process intelligent crane technology and created the "Intelligent Unmanned Crane Solutions Based on AI Dynamic Optimization Algorithms" project, setting a benchmark for the unmanned transformation of heavy-load handling in industrial scenarios. This project has overcome several key core technologies, including large-field 3D high-precision scanning, AI visual recognition, and adaptive electronic anti-sway, creating a dedicated intelligent unmanned crane solution. The overall technology is domestically leading and internationally advanced, with core technologies like scheduling control, intelligent anti-sway, and 3D scanning recognition reaching internationally leading levels. The solution can autonomously complete full-process unmanned operations such as automatic material loading/unloading and intelligent stacking, while seamlessly connecting with enterprise information systems to break information silos and eliminate data islands, solving various industry challenges in traditional warehouse operations. Backed by hardcore technology, the results are significant. After rigorous scenario validation and iterative optimization, the Keyuan intelligent unmanned crane solution achieves over 99% fully automatic warehouse operation rate and reduces on-site personnel by 90%, adapting to customized implementations in various complex warehouse environments. It has already been deployed at scale in major metallurgical enterprises such as Rizhao Steel, Shagang, and Chinalco, providing a replicable and scalable benchmark for the intelligent upgrade of the metallurgical industry.
Nanjing Qunding Technology's "AI Algorithm-Based Energy-Saving Optimization Control Platform for Data Centers" aids energy conservation and carbon reduction in the information sector. Meanwhile, the "Innovative Application of a Domestic Full-Stack Power-Specific Large Model Toolchain" project, developed by Guodian Nanjing Nanzhi Huadun Digital Technology Co., Ltd. (a subsidiary of Guodian Nanjing Automation), highlights the company's hardcore innovation in integrating digital intelligence with AI in the power industry. It also provides a replicable and scalable practical benchmark for the digital transformation of the entire power sector. The project's overall architecture includes four core modules: a model platform, an agent platform, a heterogeneous orchestration platform, and public capability components. This significantly shortens the development cycle of agent products for scenarios like power production operations, maintenance management, power trading, and general office work, providing platform support for the deployment of AI products in complex business scenarios. This domestic full-stack power large model toolchain has already been implemented and put into use in several power engineering projects. Looking ahead, the company stated it will use AI technology as a core engine to continue deepening digital R&D and scenario application, creating more benchmark demonstration projects for AI in the power sector.
Broad Service Coverage
Facing challenges such as lagging manual risk control, inefficient auditing, and blind spots in trade and global supply chains, Sumec Group, based in Nanjing for over 40 years, developed its own "AI-Powered Full-Chain Supply Chain Risk Control and Intelligent Auditing" project. By restructuring the supply chain management system with AI technology, the project has made the list. The intelligent auditing system enables pre-control and full coverage, significantly improving efficiency and accuracy in scenarios like customs declaration and document review, saving approximately 10 minutes per order. The full-chain risk control system builds a global, three-dimensional compliance control matrix with capabilities for exchange rate prediction and business opportunity mining, covering 235 countries and regions. Additionally, the intelligent assistant system integrates functions like interaction, retrieval, and document processing. An HR assistant provides 24/7 online Q&A, and key contract clauses can be extracted in just 2 minutes, comprehensively empowering efficient operations and risk supervision.
As a smart logistics technology company under China Logistics Group and a national digital logistics infrastructure platform, Zhongchu Zhihui made the list with its "AI + Integrated Logistics Solution Practice." This fully demonstrates the company's innovative achievements in applying AI to logistics and supply chain management. The solution addresses pain points in the full chain of bulk commodity road transport and multimodal transport, building a unified AI system covering cargo matching, multimodal transport network routing, path planning, intelligent customer service, intelligent pricing prediction, and a logistics credit system. On the Zhongchu Zhihui platform, when a shipper inputs transport needs, the algorithm quickly processes comprehensive data, matching vehicle types, driver credit scores, and other multi-dimensional data to quickly find suitable capacity. This solution acts like a versatile player, comprehensively supporting smart logistics. In the first five months of this year, the platform's transaction volume reached 25.66 billion yuan, a year-on-year increase of 49%; cargo volume was 285 million tons, up 40% year-on-year. The platform now serves 70,000 manufacturing, trading, and other shipper enterprises, integrating over 3.05 million road truck capacities and 33,000 ships, covering key industries like coal, steel, ore, chemicals, non-ferrous metals, grain, and new energy. Moving forward, Zhongchu Zhihui will leverage its role as a benchmark case, accelerate the promotion and replication of AI solutions, and continuously promote deep integration of AI with modern logistics, creating more digital benchmark achievements around smart logistics, green logistics, and efficient operations to help build an efficient and smooth modern circulation system.
2025 AI Application Benchmark Cases (Nanjing)
1. Guodian Nanjing Automation Co., Ltd. - "Innovative Application of a Domestic Full-Stack Power-Specific Large Model Toolchain"
2. Nanjing Keyuan Wisdom Technology Group Co., Ltd. - "Practice of Intelligent Unmanned Crane Solutions Based on AI Dynamic Optimization Algorithms"
3. Nanjing Nanzhi Advanced Optoelectronic Integration Technology Research Institute Co., Ltd. - "Photon Intelligent Engine OptoChat Assisting Photonic Chip R&D and Design"
4. Nanjing Qunding Technology Co., Ltd. - "AI Algorithm-Based Energy-Saving Optimization Control Platform for Data Centers"
5. Nanjing Fiberglass Research and Design Institute Co., Ltd. - "AI-Empowered High-Performance Glass Fiber R&D"
6. Jiangsu Sumec Group Co., Ltd. - "AI-Powered Full-Chain Supply Chain Risk Control and Intelligent Auditing"
7. Zhongchu Zhihui Technology Co., Ltd. - "Zhongchu Zhihui AI + Integrated Logistics Solution Practice"
Comments