Cap Welding Inspection
Assess low-heat battery cell welds with deep learning solutions
Graphical programming environment for deep learning-based industrial image analysis
A poorly manufactured battery cell reduces efficiency, creates an uneven load between cells that makes battery management more difficult, and decreases the lifespan of the battery pack as a whole. Errors in cell manufacture are difficult to remedy once they have been combined into modules and packs.
Once the electrodes and separator are packed into the housing of a cylindrical cell and it is filled with electrolyte, the housing is sealed by a cap. To avoid damaging the delicate electrical parts in the housing, a low-heat welding method, typically laser, is required. Such welds must be precise to ensure a secure seal around the cap. The resulting welds must be examined and passed before the cell is used inside a battery module or as a single cell. Any leakage of electrolyte through a flawed weld will lower cell efficiency and could lead to short circuits within the battery.
Proper assessment of the cap welds is key to the functionality and lifespan of the entire battery. All of these welds can vary significance in appearance, and can show a wide range of defects, but also a wide range of variation that does not affect performance. It is almost impossible to separate cosmetic from functionally significant variations with a traditional vision system because their appearances overlap.
Cognex Deep Learning defect detection and classification tools are trained on a wide range of weld variations. The system then ‘learns’ to accurately classify and distinguish different defect types despite the object and weld variations.