In the production of new energy batteries, the inspection of module weld spots is crucial. WeLinkirt's DaoAI 3D robot vision provides a reliable inspection solution for the industry with advanced technology.
User scenario: A leading new energy battery manufacturer needs to conduct high-precision inspection of the weld spots in the battery module during the weld spot inspection process on its battery module production line. The inspection objects are new energy battery modules of various specifications. The quality of the weld spots directly affects the safety and performance of the battery module.
Pain points: Traditional inspection methods have many dilemmas. Quantitatively, the miss-detection rate is as high as 3%, which causes a large number of defective products to flow into subsequent processes, increasing the rework cost and product risk. The false-alarm rate is 25%. Frequent false alarms not only waste a lot of manpower and time for re-inspection but also affect the production line rhythm. Moreover, when changing the type of battery modules of different specifications, it takes 30 minutes for manual adjustment, seriously reducing production efficiency. Combining with the hot topic of the application cases of the embodied intelligent vision perception solution of Lightelligence in robot vision and 3D grasping guidance, traditional inspection methods are obviously insufficient in terms of intelligence and automation.
Technical principle
DaoAI 3D robot vision uses a self-developed 3D camera for imaging. The 3D camera uses the principle of structured light to project a specific structured light pattern onto the inspection object. By capturing the deformation of the pattern reflected from the object surface, it can obtain the three-dimensional morphology information of the object. For weld spot inspection, it can clearly present the three-dimensional shape, height, position and other characteristics of the weld spots. In terms of algorithms, the 6D pose estimation algorithm is used to accurately determine the position and posture of the weld spots in space, achieving sub-millimeter hand-eye coordination. In this way, various defects of the weld spots, such as insufficient welding, missed welding, and welding spatter, can be accurately identified. Since three-dimensional information is richer than two-dimensional information, it can avoid misjudgment caused by factors such as reflection and shadow on the surface of the weld spots, so this technology is very effective in weld spot inspection.
- Structured light imaging obtains three-dimensional morphology and provides rich feature information.
- The 6D pose estimation algorithm accurately determines the position and posture of the weld spots.
- Sub - millimeter hand-eye coordination ensures inspection accuracy.
- Avoid misjudgment problems of reflection and shadow in two-dimensional inspection.
WeLinkirt's solution and product
Centered on DaoAI 3D robot vision, this product has the capabilities of unordered bin picking, gluing/assembly/loading and unloading guidance, and brain-eye - body closed-loop. In the inspection of new energy battery module weld spots, the self-developed 3D camera is first used to obtain the three-dimensional data of the weld spots, and then the 6D pose estimation algorithm is used for accurate analysis. At the same time, combined with the DaoAI AI AOI software system, feature recognition and semantic false-alarm filtering are carried out on the obtained data to improve the accuracy of inspection. When changing the type, the unified base of the DaoAI World model is used to realize rapid parameter adjustment and model switching, and the change-over time is shortened to 5 minutes.
DaoAI 3D robot vision provides a comprehensive and high-precision solution for the inspection of new energy battery module weld spots.
Quantitative results: Through the application of DaoAI 3D robot vision, the detection rate of weld spots has been increased to 99.2%, and the miss-detection rate has been reduced to <0.8%, effectively avoiding the outflow of defective products. The false-alarm rate has been reduced by -60%, greatly reducing the manpower and time cost of re-inspection. The change-over time has been shortened from 30 minutes to 5 minutes, significantly improving the change-over efficiency of the production line and meeting the production needs of battery modules of different specifications.
FAQ
How does DaoAI 3D robot vision improve the accuracy of weld spot inspection?
The self-developed 3D camera is used to obtain the three-dimensional morphology of the weld spots. The 6D pose estimation algorithm is used for accurate analysis. Combined with the feature recognition and false-alarm filtering of the AI AOI software system, it avoids misjudgment in two-dimensional inspection and improves accuracy.
How much time can be saved during model change with this product?
The traditional method takes 30 minutes for model change. Using DaoAI 3D robot vision combined with the DaoAI World model, the model change time can be shortened to 5 minutes.
By how much has the false-alarm rate been reduced after applying this product?
After application, the false-alarm rate has been reduced by -60%, greatly reducing the manpower and time cost of re-inspection.