Remarkable Results of AI Vision Inspection for Defects in Electronic FPC Flexible Boards

WeLinkirt's advanced AI vision technology has brought new breakthroughs in defect inspection of electronic FPC flexible boards, effectively solving many problems of traditional inspection methods.

In today's booming electronics industry, FPC (Flexible Printed Circuit) flexible boards are indispensable connecting components in electronic products. The quality of FPC flexible boards directly affects the performance and stability of the entire product. Take the FPC flexible board production line of a leading electronics manufacturer as an example. After the assembly process, comprehensive and accurate defect inspection of FPC flexible boards has become a crucial step to ensure product quality. FPC flexible boards are widely used in various electronic products such as mobile phones, computers, and smart wearable devices. Quality problems of FPC flexible boards may lead to unstable signal transmission, device failures and other serious consequences. Therefore, defect inspection of FPC flexible boards is particularly important.

Pain Points: Why Is It Difficult?

Traditional defect inspection methods for FPC flexible boards face many quantitative dilemmas. In terms of the missed detection rate, about 3% of missed detection means that some defective products may flow into the market. Once these defective products enter the consumer market, they will have a serious impact on the product reputation, leading to a decrease in customer satisfaction and even potential after-sales disputes and increased recall costs. For example, if a circuit break in the flexible board is not detected, the electronic product may experience malfunctions such as freezing and signal interruption during use.

The false alarm rate is also a major pain point of traditional inspection methods. A false alarm rate as high as 20% means that a large number of qualified products are misjudged as defective. This not only increases the re-inspection workload but also raises the production cost. Each misjudgment requires additional manpower and time for re-inspection, resulting in a waste of production resources. Moreover, frequent re-inspections will also extend the production cycle and reduce production efficiency.

The complex texture of FPC flexible boards is also a difficult problem for traditional inspection methods to overcome. The texture structure of flexible boards is diverse, and there may be slight differences between different batches of products. This makes it difficult for traditional methods to establish a unified and effective inspection standard. For defects that have never been seen before, traditional methods lack effective detection means. Because traditional inspections often rely on preset rules and templates, they cannot accurately identify new types of defects.

The long change-over time is also an important factor restricting production efficiency. Traditional methods require 30 minutes for each change-over, which is a considerable time cost in large-scale production. Frequent change-over operations will cause the production line to stop and restart, affecting the continuity and stability of production and reducing the overall production efficiency.

Technical Principle

WeLinkirt uses an unsupervised anomaly detection algorithm combined with advanced imaging technology and hardware equipment to solve the above problems. The unique feature of the unsupervised anomaly detection algorithm is that it only needs to model good products. By learning the characteristics of a large number of good products, the algorithm can identify abnormal areas that are different from the characteristics of good products, thereby detecting defects that have never been seen before. Normal FPC flexible boards have a certain texture and feature distribution. When a defect occurs, these textures and features will change. The algorithm establishes a normal feature model through learning from a large number of good products. When the detected sample has a large difference from the model, it is judged as a defect. Compared with traditional methods, this algorithm does not require preset complex rules and templates, and can automatically adapt to different products and defect types, greatly improving the flexibility and accuracy of detection.

Advanced imaging technology is an important part of WeLinkirt's inspection system. It can clearly capture the surface details of FPC flexible boards and provide accurate data for the algorithm. Through high-resolution imaging equipment, even tiny scratches and stains can be clearly recorded. The hardware equipment has high-precision detection capabilities and can detect defects at the micron level. This enables WeLinkirt's inspection system to detect subtle defects that are difficult to detect by traditional methods, improving the detection accuracy. Moreover, the algorithm has self-learning ability and can continuously learn from the feedback of the production line to continuously optimize the detection effect. With the continuous accumulation of data during the production process, the algorithm can gradually adapt to new defect types and product changes, further enhancing the accuracy and reliability of detection.

Typical Application Scenarios

• Scratch detection: Scratches on the surface of FPC flexible boards may affect the conductivity of the circuit and cause signal transmission problems. WeLinkirt's inspection system captures images of the flexible board surface through imaging technology, and the algorithm analyzes the color and texture changes in the image to identify the location and size of the scratches. The difficulty lies in that scratches may be very subtle, and high-resolution imaging equipment and precise algorithms are required for accurate identification.
• Stain detection: Stains may interfere with signal transmission or affect the appearance quality of the flexible board. The detection system detects stains by analyzing abnormal color areas in the image. Since the colors and shapes of stains vary and may be similar to the normal texture of the flexible board, accurately distinguishing stains from normal textures is the difficulty of detection.
• Circuit break detection: Circuit break is one of the most serious defects in FPC flexible boards, which will cause electronic products to malfunction. WeLinkirt uses imaging technology to obtain images of the circuit, and the algorithm analyzes the continuity and characteristics of the circuit to determine whether there is a break. The complex layout and small size of the circuit increase the difficulty of detection, and high-precision hardware equipment and advanced algorithms are required to ensure the accuracy of detection.
• Hidden solder joint detection: The quality of hidden solder joints directly affects the connection stability of the flexible board. The self-developed 3D camera of the DaoAI 2D / 3D AI AOI device can perform three-dimensional morphology reconstruction to detect the location, shape and welding quality of hidden solder joints. Since hidden solder joints are located inside the flexible board, they are difficult to detect by traditional methods, while 3D reconstruction technology can provide more comprehensive information and improve the accuracy of detection.
• Coplanarity detection: Coplanarity refers to the flatness of the flexible board surface. Failure to meet the coplanarity requirements may affect the assembly of the flexible board with other components. The detection system measures the height change of the flexible board surface through 3D imaging technology to determine whether the coplanarity meets the standard. The flexibility and complex shape of the flexible board make coplanarity detection challenging, and precise measurement and analysis methods are required.

Implementation Case

A leading electronics manufacturer, which occupies an important position in the global electronics market, has a large-scale FPC flexible board production line with a daily output of thousands of pieces. Before introducing WeLinkirt's AI vision inspection solution, the manufacturer had been facing many problems caused by traditional inspection methods, such as high missed detection rate, high false alarm rate, and long change-over time. During the implementation process, WeLinkirt's technical team closely cooperated with the manufacturer's production department to conduct a comprehensive evaluation and transformation of the production line. First, the DaoAI AI AOI software system and the DaoAI 2D / 3D AI AOI device were installed, and the system was debugged and optimized to ensure its smooth connection with the existing production line. Then, the operators were professionally trained to familiarize them with the operation and maintenance of the system.

WeLinkirt's AI vision inspection technology has brought revolutionary changes to the defect inspection of FPC flexible boards, significantly improving the inspection efficiency and accuracy.

Before the implementation, the missed detection rate of defects in the manufacturer's FPC flexible boards was about 3%, the false alarm rate reached 20%, and the change-over time for each operation was 30 minutes. After the implementation, the defect detection rate increased to 98%, the missed detection rate decreased to <2%, the false alarm rate decreased by -70%, and the change-over time was shortened to 5 minutes. These significant changes in data indicate that WeLinkirt's solution effectively solves the pain points of the manufacturer in FPC flexible board inspection, and improves the production efficiency and product quality.

WeLinkirt's Solution and Products

WeLinkirt provides the DaoAI AI AOI software system and the DaoAI 2D / 3D AI AOI device. The DaoAI AI AOI software system uses the feature recognition of the visual basic model. With only one good product, 0-code automatic programming can be completed in 5 minutes. Using APDT positive-sample/few-sample learning (only 1-20 good products are required), the time cost of sample collection and programming is greatly reduced. At the same time, the software also has a semantic false alarm filtering function, which can effectively reduce the false alarm rate and improve the accuracy of detection.

The self-developed 3D camera of the DaoAI 2D / 3D AI AOI device can perform three-dimensional morphology reconstruction, which can detect hidden solder joints, coplanarity and micron-level morphology. Three - dimensional morphology reconstruction technology provides more abundant information for detection, enabling the detection system to find hidden defects that are difficult to detect by traditional methods, and greatly improving the accuracy of detection. In the implementation process, the software system is combined with the hardware device to achieve efficient defect detection. The software system is responsible for analyzing and processing images, and the hardware device is responsible for collecting images and performing detection operations. The two cooperate with each other to provide a comprehensive solution for FPC flexible board defect detection.

Quantitative Results

By using WeLinkirt's solution, the manufacturer has achieved remarkable results in FPC flexible board defect inspection. The defect detection rate has increased to 98%, which means that more defects can be detected and processed in time, effectively reducing the risk of defective products flowing into the market. The missed detection rate has decreased to <2%, a significant drop compared with before the implementation. The false alarm rate has decreased by -70%, greatly reducing the re-inspection workload and improving production efficiency. The change-over time has been shortened from 30 minutes to 5 minutes, and the continuity and stability of the production line have been significantly improved, further enhancing the overall production efficiency. These quantitative results fully demonstrate the excellent performance and great value of WeLinkirt's AI vision technology in FPC flexible board defect inspection.

This article was generated by AI. Customer cases are simulated scenarios based on real product capabilities and figures are illustrative; see product pages for official benchmarks.