The Jiangsu Provincial Department of Industry and Information Technology recently announced the proposed arrangements for the third batch of the 2026 Jiangsu Provincial Special Funds for Building a Strong Manufacturing Province, covering intelligent transformation, digital transformation, and networked connectivity projects. Two enterprises from Wuzhong District made the list.
Suzhou Collaborative Innovation Intelligent Manufacturing Technology Co., Ltd. was selected for an industrial large model construction project. Suzhou Luster LightTech Industrial Intelligent Technology Co., Ltd. was selected for a high-quality dataset construction project in key industries.
Project Brief: R&D and Industrial Application of an Industrial Large Model Based on Multimodal Fusion and Identification Resolution
The project focuses on the research, development, and industrial application of an industrial large model based on multimodal fusion and identification resolution. It innovatively builds a "1+1+3+N+1" industrial large model industrialization application architecture and develops the self-controllable "Gongdao Large Model." Targeting Jiangsu's three major advantageous industrial chains—electronic information, high-end equipment, and textile and chemical fiber—the project will develop no fewer than 50 industrial intelligent agents, covering scenarios such as intelligent quality inspection, predictive equipment maintenance, production line scheduling, and energy consumption optimization, and will create three industry demonstration projects. Once completed, the project will break partial dependence on foreign industrial intelligent technologies, lower the threshold for enterprise digital transformation, and, through platform-based and ecosystem-oriented promotion, continue to provide strong support for more manufacturing enterprises to improve quality and efficiency and achieve transformation and upgrading.
Project Brief: Construction and Application of a High-Quality Dataset for New Energy Industrial Vision
Led by Suzhou Luster LightTech and jointly built with CAICT (Jiangsu), Wuxi Enjie, and Ningde Zhuogao, this project focuses on key scenarios such as lithium battery separators, electrode sheets, battery cells, and photovoltaic glass. It aims to build multimodal data resources covering 50 million industrial images, 5 million fine annotations, and more than 40 defect types, while simultaneously linking equipment, process, and quality parameters, enabling data to move from "seeing defects" to "explaining causes." With a "4 libraries + 1 platform" structure at its core and relying on a trusted data space to achieve cross-enterprise circulation that is "usable but invisible," the project forms a closed loop of "technology supply + standards leadership + scenario verification." It can empower more than five industry models and more than three quality inspection intelligent agents, achieving defect classification accuracy of over 99.7% and yield improvement of more than 0.5%, providing a replicable and promotable high-quality data foundation for new energy intelligent manufacturing.
Comments