On March 11, 2026, at the Guangzhou Nansha Technology Innovation and Industrial Innovation Integrated Development Conference, Zhu Wu, Secretary-General of the Nansha District High-tech Enterprise Association, released Nansha's first batch of Innovation Opportunity List. The list is built around the concepts of "demand-driven, scenario-driven, industry-research integration, and market implementation," constructing a full-chain innovation ecosystem from R&D to market.
Two products from our team, the "Automated Surface Defect Inspection System for Industrial Components" and the "Software Engineering Documentation Agent," were selected for the scenario application list.
This product serves high-end manufacturing sectors such as aerospace, gas turbines, and shipbuilding. It performs automated high-precision defect detection, defect dimension measurement, and workpiece qualification assessment on heterogeneous components, addressing the limitations of existing industrial vision inspection products, including low versatility, incomplete detection coverage, and insufficient detection capability. The system integrates key technologies including 2D + 3D data acquisition, multi-dimensional data fusion, robotic control, and collaborative operation control. It can detect and identify defects as small as 0.1mm on industrial components ranging from 0.1 to 1 meter in size, with comprehensive defect type coverage and an accuracy rate exceeding 95%, helping customers reduce costs, improve quality, and enhance efficiency.
This product integrates cutting-edge technologies such as computer vision and large language models. Leveraging extensive software requirements specification documents (including standards, work instructions, requirements, proposals, designs, contracts) and test case data covering 8 quality characteristics including functionality, performance, compatibility, and security, it trains specialized document generation models through advanced reinforcement learning and modeling techniques. The system enables automated intelligent generation of complete documentation throughout the software lifecycle, including plans, test cases, records, and reports, addressing the core pain points of traditional manual documentation: low efficiency, inconsistent quality, and difficulty in meeting strict compliance requirements.
It is reported that this list comprehensively covers key areas such as AI and intelligent manufacturing, providing various innovation entities with a clear path from technology development and scenario validation to market implementation, helping Nansha accelerate its development as a hub for deep integration of technological and industrial innovation in the Guangdong-Hong Kong-Macao Greater Bay Area.