On August 28th, the 19th Frost & Sullivan Global Growth, Innovation and Leadership Summit and 4th New Investment Conference (hereinafter referred to as '2025 Frost & Sullivan New Investment Conference') AI Evolution Theory - Physical World Intelligent Systems sub-forum was held in Shanghai, hosted by Frost & Sullivan, a globally leading growth consulting company (Frost & Sullivan, abbreviated as 'Frost & Sullivan').
At this forum, Dr. Liu Jun, Chairman of Rilian Technology, shared a keynote speech titled "AI Leading the Transformation of the Industrial X-ray Inspection Industry".
Dr. Liu Jun, Chairman of Rilian Technology
The following are the key points from Dr. Liu Jun's speech:
AI is reshaping the industrial inspection landscape
Dr. Liu Jun pointed out that AI technology innovation leads a new era of industrial radiographic inspection, deeply empowering the industry's rapid development. The prominent features are:
1. See more clearly: AI automatically enhances images to identify subtle defects (such as micron-level cracks, bubbles, impurities).
2. Faster detection: By using pre-trained models for millisecond-level real-time analysis and classification of X-ray images, automatic determination of quality is achieved, significantly shortening the detection cycle.
3. More accurate judgment: AI models trained on massive defect samples can continuously learn and optimize their judgment criteria, effectively reducing the error rate and false negative rate.
4. Learn more deeply: AI can analyze potential patterns and root causes of defects through data mining, providing data-driven decision support for process optimization and quality control.
At the same time, Dr. Liu Jun believes that the core of AI applications in industrial radiographic inspection is image data, which places higher demands on enterprise data collection and management. Market demand forces enterprises to develop more intelligent industrial inspection equipment, and enterprises use AI technology applications to improve inspection efficiency and accuracy.
Rilian Technology's Layout in the AI Field
Dr. Liu Jun pointed out that AI technology has become one of the core technologies, establishing Rilian Technology's leading position in the field of intelligent detection. Rilian Technology has built an AI technology platform, connecting the channel from algorithm research to terminal applications. Rilian Technology's positioning of its AI technology platform is strategic: it is the core hub for technology asset accumulation, the core platform for AI development and resource management, the core carrier for multi-industry applications, and the technical positioning is also clear: it is a platform for technical assets, AI development, and resource management.
Rilian Technology has been conducting AI research since 2013. After more than a decade of development in AI technology, it has accumulated a vast amount of real industrial inspection scenario data. AI technology has become one of Rilian Technology's core competitiveness. Rilian Technology deeply integrates and makes forward-looking arrangements for the three cornerstones of AI - "computing power - algorithms - computing materials", and constructs a solid triangular architecture for AI R & D and application. In terms of computing power, Rilian Technology is making every effort to build the first large-scale industrial X-ray imaging data center and dedicated computing power center in China at its Wuxi headquarters, further strengthening the company's strength in AI hardware. In terms of algorithms, it has established a Shanghai Artificial Intelligence R & D Center. Relying on a high-level and professional AI algorithm and large model R & D team, it has delved deeply into the field of industrial intelligence and constructed a unique and complete industrial artificial intelligence technology system. In terms of computing materials, Rilian Research Institute deeply processes and controls industrial image data through unified aggregation, standardized cleaning, deep annotation and security management. At present, it has built a leading industrial inspection dedicated database with tens of millions of records in the industry.
Relying on its long-term accumulation in AI, Rilian Technology has successfully completed its strategic transformation from a high-end equipment manufacturer to an AI-driven industrial inspection solution provider, and is gradually becoming a global leader in industrial X-ray inspection equipment.
In terms of technology, RelyonSoft has made breakthroughs in core algorithm areas, building a strong moat. The main breakthroughs are as follows:
Complex structure detection: Develop advanced deep learning models for multi-layer PCBs, battery interiors, complex castings, and other structures, significantly improving the detection rate and classification accuracy of tiny and concealed defects.
Lack of samples: Innovating application of small sample learning and transfer learning techniques effectively solves the problem of rare defective samples in industrial scenarios;
Improve computing efficiency: Continuously optimize model efficiency, significantly reduce computing load while ensuring ultra-high precision, and enable algorithms to efficiently adapt to various edge computing devices and production line terminals.
In terms of large models, RAIN Technologies has developed its own AI vertical large model and released an original automatic defect recognition (ADR) enhancement application.
Dr. Liu Jun pointed out that industrial inspection equipment paired with AI enjoys a good reputation among clients. The sufficient customer base and AI reserves have enabled Rilian Technology to complete its strategic transformation from a high-end equipment manufacturer to an AI-driven industrial inspection solution provider.
