With the rapid popularity of new energy vehicles and the continued expansion of the power battery industry, the production standards and quality and safety control of lithium batteries are becoming increasingly stringent and mature. The traditional 2D vision-based inspection solution can no longer meet the increasingly demanding inspection requirements of power batteries, and the limitations of traditional algorithms for defect detection are also hindering the digitalization of the power battery industry.

For the small, diverse, varied and often with concave and convex marks nature of power battery defects, Aqrose Technology, together with LMI, an internationally renowned 3D vision manufacturer, launched a solution of high-precision 3D line laser camera with AI vision inspection platform, providing an excellent solution for power battery defect detection.

Aqrose Technology's AI vision inspection platform software (AIDI) integrates labeling, training and model tuning to solve the problem of locating, detecting and classifying complex defects without programming, and is suitable for various complex defect detection scenarios with strong compatibility. At present, AIDI has released a powerful 3D online inspection feature, fully supporting mainstream 3D format data input, with the ability to perform all-round AI+3D inspection of power batteries, together with LMI 3D line laser camera, providing an all-round intelligent solution for quality control of power batteries.

I. Detection of defects on the surface of the blue membrane of the Battery cell

In the production of lithium battery cells, an important process is to wrap the blue membrane to achieve the effect of dust and water resistance. However, bubbles and poor cutting often occur during the lamination process, and damage from collisions and friction can also occur during the circulation process, so the appearance inspection of the blue membrane surface of the battery cell is an important part of the quality check. Traditional inspection solutions can not distinguish and detect defects with weak color contrast such as scratches, dirt, bubbles and pits, which can be addressed by using 3D imaging with AI vision algorithms.

II. Battery cell top cover welding defect detection

After the bare core is put in the housing, the cap is closed and soldered shut. After welding, the quality of the weld is tested and controlled in order to avoid sealing problems caused by poor welding. As the weld bead is at the root edge, the conventional way of acquiring images does not work. By using 3D line laser tilt scanning, Aqrose Technology is able to obtain a clear 3D image of the weld bead, and then uses an AI vision inspection platform to precisely pinpoint the defects, providing an excellent solution to the problem of welding defects such as bursts, bumps, polarization, pinhole that cannot be well detected by conventional vision.

III. Fill port welding appearance inspection

In the general process, after being filled with electrolyte, the housing is then encapsulated with fill port welding. If the welding is poor and the electrolyte leaks out, safety accidents can easily occur, so the quality inspection of fill port welding is an important part of ensuring battery safety. With conventional 2D image and inspection solutions, it is difficult to distinguish between good and bad welding quality. The problem can be well solved by acquiring the high and low segment difference data with 3D line laser, combining it with Aqrose Technology's AIDI to generate a 3D heat map, and then using AI inspection.

Aqrose technology was founded in 2017, and its core team originates from Tsinghua University's AI Lab, focusing on applying leading AI and machine vision technologies to the industrial sector to empower China's most outstanding advanced manufacturing and smart logistics customers. The AI industrial vision platform software AIDI, edge-end software and hardware integrated series products Ebox, vertical industry AI solutions launched have been widely used in consumer electronics, automotive, new energy, pharmaceutical, semiconductor, logistics and other industries, and more than a hundred factories have deployed online AIDI, including dozens of benchmark companies in advanced manufacturing and intelligent logistics, which has become the leader of industrial AI platforms in China.