Aijiang Technology: Forging Ahead in Intelligent Inspection Equipment to Support Domestic Substitution of High‑End Manufacturing-Reported by China Youth Network

Against the backdrop of a trillion‑dollar‑scale market, intelligent inspection equipment is accelerating domestic substitution, bringing high‑end manufacturing quality control toward a traceable closed‑loop system.

China Youth Network News — Driven by the iterative upgrading of domestic high‑end manufacturing toward greater precision, intelligence and autonomy, industrial quality inspection, as a core link for product quality control, production‑process optimization and industrial‑chain security reinforcement, is undergoing structural transformation. The intelligent‑inspection industry is therefore entering a phase of rapid growth.

Acting as the “neural perception system” for high‑end manufacturing, intelligent inspection builds upon industrial sensors and instruments by integrating perception, analysis and feedback capabilities. It can capture micron‑level product deviations with high precision — a feat unattainable by conventional manual inspection or single‑mode mechanical detection.

For a long time, China’s high‑end precision‑inspection sector has been plagued by high technical barriers, heavy reliance on imports, insufficient traceability and obstacles to industrial application. To tackle these pain points, many local enterprises are advancing breakthroughs in technical bottlenecks through industry‑university‑research collaboration, lifting the overall competence of China’s domestic intelligent‑inspection industrial chain.

In terms of home‑grown core components, enterprises are improving localization rates via independent R&D, cooperation with upstream suppliers, and innovation based on imported technologies. They are accelerating the adoption of domestic components and algorithm platforms in high‑end inspection equipment. For instance, Uni X‑ray has achieved technological breakthroughs and advanced industrialization for micro‑focus X‑ray sources; Hikrobot and other players have enhanced their industrial‑vision hardware and algorithm platforms through full‑stack independent development.

While the installed base of domestic inspection equipment has expanded noticeably, weaknesses persist in high‑end segments. Data released by the State Administration for Market Regulation shows that by the end of 2024, domestically‑produced instruments accounted for over 93 % of the total quantity of equipment owned by Chinese inspection‑and‑testing institutions, and approximately 65.18 % in terms of original asset value. This demonstrates markedly strengthened indigenous R&D capacity for core inspection equipment. Nevertheless, gaps remain: there is substantial room to raise the localization rate of high‑end industrial cameras and key optical components, as well as metrology‑and‑inspection equipment for the front‑end semiconductor manufacturing process.

Industry experts note that intelligent‑inspection capability depends not merely on hardware performance, but also on generic‑technology platforms, scenario adaptation and deep integration with manufacturing workflows. Ouyang Jingsong, Deputy Director of the Expert Advisory Committee for the Intelligent Inspection Equipment Industry and Director of the Instrumentation Technology & Economy Institute under China Machinery Industry Federation, argues that intelligent‑inspection equipment must be closely tailored to real‑world scenarios and advance in tandem with related industries. Tight integration between inspection hardware and manufacturing scenarios will drive equipment toward higher precision, reliability and environmental adaptability.

In response to this trend, enterprises are developing vertical‑industry solutions for semiconductors, display panels, new‑energy industries, aerospace and other sectors. They build process databases and defect‑sample libraries to boost inspection‑system adaptability to specific materials, production steps and factory‑floor conditions. In the niche field of aerospace power‑supply inspection, Wuhan Aijiang Technology Co., Ltd. (“Aijiang Technology”) ranks among the early domestic enterprises to realize equipment localization. Its extra‑large‑aperture AM0 (extraterrestrial standard solar spectrum) solar‑simulator serves as China’s first dedicated inspection platform for ultra‑large solar‑cell arrays.

A Hubei‑provincial “Specialized, Refined, Differential and Innovative” SME and a Wuhan‑level above‑scale industrial enterprise, Aijiang Technology has operated for more than a decade. It delivers inspection equipment and system solutions covering photovoltaic, energy‑storage, semiconductor and aerospace applications and holds over one hundred patents. Starting in 2019, the company targeted technical challenges including spectral calibration, large‑aperture light‑source uniformity and ultra‑high‑power power‑supply stability. Through algorithm optimization and mechanical‑structure iteration, it developed domestically‑built inspection hardware for aerospace solar‑cell arrays, achieving an overall hardware‑and‑software localization rate exceeding 92 %.

Aijiang Technology has also invested in perovskite‑cell inspection. It operates a specialized testing laboratory and has built a standardized dataset covering 25 categories of perovskite materials and 38 types of quantum‑dot structures. Its relevant inspection products have been listed in the Provincial Catalogue of Innovative Products.

“Conventional inspection systems rely heavily on senior engineers to trace defect root causes, identify process‑step origins and optimize parameters. Such workflows are time‑consuming and subject to human experience,” says Yuan Chi, Deputy General Manager of Aijiang Technology. As artificial‑intelligence technologies evolve toward cognitive intelligence, high‑end manufacturing demands AI capabilities beyond isolated tools: inspection workflows require end‑to‑end collaboration, extending from defect detection to process tracing and parameter tuning.

In 2025, Aijiang Technology launched its self‑developed Xinghan AI Multimodal Vertical Large Model. Trained on massive volumes of industrial‑scene data and empowered by proprietary process knowledge graphs, the model adopts a Mixture‑of‑Experts (MoE) architecture and bidirectional‑encoder technology. It fuses multi‑dimensional inputs including images, spectral data, 3D imaging, electrical‑performance metrics and textual information. “In selected scenarios, the system reaches a defect‑recognition accuracy of 98.7 %, boosts inspection efficiency by more than 12‑fold and cuts overall inspection costs by roughly 40 %,” Yuan Chi explains.

Professor Liu Kan, Dean of the School of Computer Science and Artificial Intelligence at Wuhan Textile University, which co‑established the Optoelectronic‑Intelligence Joint Research Center with Aijiang Technology, observes that future competition in China’s intelligent‑inspection industry will hinge not on individual algorithms or hardware alone. Success will rest upon accumulated domain‑process expertise, multimodal‑data‑fusion competence and localized field‑validation capacity. High‑end inspection applications impose strict requirements for stability, traceability and data security. For domestic equipment to move from “functional” to “high‑performing”, sustained progress is needed in standard‑system construction, production‑line adaptation and long‑term real‑world verification.

In the sphere of industry‑university‑research cooperation, Aijiang Technology has co‑founded the Optoelectronic‑Intelligence Joint Research Center with Wuhan Textile University and the Wuhan Advanced‑Technology Research Institute. It has participated in key municipal R&D initiatives for artificial intelligence and been selected as a typical AI‑application scenario in Hubei Province. Following a collaborative framework where “enterprises raise real‑world problems, universities conduct research, both sides jointly deliver solutions, and outcomes are translated into industrial products”, the company is advancing R&D and commercialization of large‑area perovskite‑cell inspection equipment.

Corporate practice shows that domestic‑substitution efforts are shifting from pure hardware replacement toward comprehensive system‑capacity building. Domestic intelligent‑inspection equipment is progressing from “workable” to “reliable, preferred and trusted”. Industry professionals point out that high‑end localization entails far more than hardware retrofitting; it calls for long‑term accumulation of process know‑how, data standards and engineering‑validation experience.

Regulators have aligned with this vision. The State Administration for Market Regulation plans to organize digital‑technology R&D and demonstration projects for inspection‑and‑testing, piloting digital‑intelligent workflows in eligible industries. Experts anticipate that deeper integration between AI and manufacturing will expand intelligent inspection from discrete quality‑control tasks into foundational capabilities spanning design, production, quality assurance and process iteration. Realizing this vision will demand continuous advances in core‑component localization, vertical‑scenario adaptation and closed‑loop data capabilities.

(Reported by Liu Hongkai, Hu Mengzhuo)

Domestic substitution means more than hardware replacement. It represents long‑term accumulation of process knowledge, closed‑loop data workflows and engineering validation.

Going forward, Aijiang Technology will remain focused on photovoltaic, energy‑storage, aerospace and power‑semiconductor sectors. It will keep refining intelligent‑inspection hardware, advance multimodal AI‑driven inspection technologies and deepen industry‑university‑research partnerships. With robust proprietary technologies, the company aims to facilitate high‑end‑manufacturing upgrading and help domestic inspection equipment earn the reputation of being reliable, preferred and trusted.

Industry peers are welcome for exchanges and cooperation.