In the electronics/PCBA industry, the micro - crack detection of passive components like MLCC is of great importance. Traditional detection methods have many problems, and AI vision technology provides an effective way to solve these problems.
User scenario: On the production line of a leading electronics/PCBA manufacturer, after the assembly process of passive components like MLCC, micro - crack detection of the components is required. MLCC, as a common passive component, is widely used in various electronic products. The existence of micro - cracks may affect the performance and stability of the products. Therefore, the detection of micro - cracks is a key link to ensure product quality.
Pain points: Traditional detection methods mainly rely on sampling inspection and cannot achieve 100% full inspection, resulting in a relatively high miss - detection rate. According to statistics, the miss - detection rate can reach about 3%. At the same time, the false - alarm rate of manual detection is also relatively high, about 15%. This not only increases labor costs but also reduces production efficiency. Moreover, when changing product models, the programming and debugging time of traditional detection equipment is relatively long, generally more than 30 minutes, which seriously affects the flexibility of the production line and the production rhythm. In addition, with the improvement of industry quality standards, traditional detection methods are difficult to meet compliance requirements.
Technical principle
WeLinkirt's AI vision technology uses advanced algorithms and imaging principles to solve the micro - crack detection problem. In terms of algorithms, deep learning algorithms are used. Through a large amount of sample data for training, the model can accurately identify the features of micro - cracks. Deep learning algorithms have powerful feature extraction and classification capabilities, which can extract the subtle features of micro - cracks from complex images, thus achieving high - precision detection. In terms of imaging, a self - developed 3D camera is used for image acquisition. The 3D camera can obtain the three - dimensional morphology information of the components. Compared with traditional 2D imaging, it can more comprehensively show the surface conditions of the components, and can also clearly image the micro - cracks hidden inside or on the side of the components. This three - dimensional morphology reconstruction technology enables WeLinkirt's AI vision system to detect micron - level micro - cracks, greatly improving the accuracy and reliability of detection.
- Deep learning algorithms can automatically learn the features of micro - cracks, avoiding the limitations of traditional algorithms that require manual design of features.
- The use of 3D cameras provides more abundant image information, which helps to improve the accuracy of detection.
- Three - dimensional morphology reconstruction technology can accurately restore the surface information of components, providing a more reliable basis for micro - crack detection.
- Through continuous model training and optimization, the system can adapt to the detection needs of different types and specifications of MLCC components.
WeLinkirt's solutions and products
WeLinkirt provides the DaoAI AI AOI software system and the DaoAI 2D / 3D AI AOI equipment to solve the micro - crack detection problem of passive components like MLCC. The DaoAI AI AOI software system has powerful feature recognition capabilities. Based on the visual basic model, it can achieve 0 - code automatic programming for a good product in 5 minutes. Using the APDT positive - sample/few - sample learning technology, only 1 - 20 good - product samples are needed to complete model training, greatly shortening the model training time. At the same time, the software system also has a semantic false - alarm filtering function, which can effectively reduce the false - alarm rate. The DaoAI 2D / 3D AI AOI equipment has a self - developed 3D camera. Combined with the three - dimensional morphology reconstruction technology, it can detect hidden solder joints, coplanarity, and micron - level morphology, providing a more comprehensive solution for micro - crack detection. During the implementation, WeLinkirt integrates the software system and the equipment, deploys them through SDK / API / Docker, and supports 100% local privatization to ensure that the data does not leave the factory, guaranteeing the data security of customers.
WeLinkirt's AI vision technology provides an efficient, accurate, and secure solution for the micro - crack detection of passive components like MLCC.
Quantitative results: By using WeLinkirt's solutions, the detection rate of micro - cracks in passive components like MLCC of the leading electronics/PCBA manufacturer has reached over 98%, and the miss - detection rate has been reduced to <2%. The false - alarm rate has been reduced by - 70%, greatly reducing the workload of manual re - inspection. The product model change time has been shortened from more than 30 minutes to 5 minutes, improving the flexibility of the production line and production efficiency.
FAQ
What are the problems with traditional micro-crack detection methods for passive components such as MLCC?
Traditional methods rely on sampling inspection and can't conduct 100% full inspection. The missed-detection rate is about 3%, and the false-alarm rate of manual inspection is about 15%, increasing labor costs and reducing efficiency. Programming and debugging for product changeover take over 30 minutes, and it's hard to meet industry quality standards.
What is the principle of DaoAI's AI vision technology?
DaoAI uses deep-learning algorithms trained with a large number of samples to accurately identify micro-crack features. It also uses self-developed 3D cameras to collect images, obtain 3D morphology information, detect micron-level cracks, and improve detection accuracy and reliability.
What solutions and products does DaoAI provide?
DaoAI provides the DaoAI AI AOI software system and DaoAI 2D / 3D AI AOI equipment. The software can perform zero-code automatic programming, shorten model training time and reduce false-alarms. The equipment combines 3D technology to provide a comprehensive detection solution.