Lithium ion machine vision market is heating up

Demand for machine vision increases

With the large-scale manufacturing of power batteries, the demand for machine vision in the entire process chain including Top 10 power battery companies has increased significantly. In 2021, China’s power battery shipments will be 220GWh, and it is expected to reach 450GWh in 2022. The market size of machine vision will also increase from 880 million to 1.1 billion last year to 1.8 billion to 2.25 billion RMB.

Square winding cell process

It can be seen that the production process of lithium-ion batteries is complex, involving many processes, and different processes require different machine vision inspection systems.

Taking a square wound cell as an example, in the front-end process of the cell, visual inspection is mainly used for coating deviation correction, size measurement, surface defect detection of pole pieces, size measurement, winding alignment, etc.

In the post-cell process, visual inspection is mainly used in the inspection of the tab folding of the bare cell, the inspection of the tab cutting debris, the inspection of the tab welding quality, the size measurement, and the positioning of the glue. As well as the welding quality inspection of the top cover of the shell, the welding quality inspection of the sealing nails, the appearance inspection of the cell, etc.

At the same time, in the module and PACK section, visual inspection equipment is mainly used, which is mainly used in bottom blue glue inspection, BUSBAR weld inspection, side weld inspection, module full-size inspection, PACK inspection, etc.

It is worth mentioning that with the improvement and refinement of the measurement requirements of cells, modules and PACKs, the conditions of the measured objects are becoming more and more complex.

3D vision can solve the defect that traditional 2D vision cannot accurately detect high features, and has low dependence on the external environment, and has become an important part of lithium-ion machine vision.

Machine vision system industry chain

From the perspective of the market competition pattern, since the machine vision system consists of core software and hardware such as light sources, lenses, industrial cameras, and visual control systems, there are many participants, and some companies have even expanded to the entire industry chain. Therefore, the leading companies in each link are also not exactly.

At the same time, due to the large differences in the demand for machine vision in different industries and different scenarios. The understanding of the lithium battery manufacturing process, the accumulation of a lot of experience, repeated trials, and precipitation of know how have become one of the core barriers to entering this field. At present, a number of manufacturers with high market recognition have been formed in the field of lithium-ion machine vision.

Machine vision system integrators mainly include: Yihong Intelligent, Supersonic, APT, Lingyunguang, Xinshizhike, Samson, Huahan Weiye, Su Yingshi, Simou Technology, Huashi Intelligent, etc.

Light source suppliers mainly include: CCS, OPT, etc.

Lens suppliers mainly include: Opto, Hikvision, Huarui, Long Trail, Dongzheng, Morit, Schneider, etc.

2D industrial camera manufacturers include: OPTE, Allied Vision, Hikvision, Keyence, Sik, Hua Rui, DALSA, Cognex, etc.

3D vision sensor/3D camera manufacturers include: LMI, Keyence, Optim, Deep Vision, Mecamand, Sick, Cognex, smartray, etc.

Among them, in the pre-cell process, system integrators such as Yihong Intelligent, Opto, Supersonic, Xinshizhike, Lingyunguang, and Huashi Intelligent performed well in the market; In the post-production process of the cell, Opte, Yi Hong, Huahan Weiye, Angshi, Su Yingshi, etc. performed even more prominently.

In the module PACK visual inspection equipment, Yihong Intelligent, Supersonic, Su Yingshi, etc. have taken the lead in the market.

Lithium ion machine vision market

Full-line visual inspection

Full-line visual inspection will become standard for power battery companies, power battery companies have accelerated production expansion, becoming a continuous growth point for machine vision companies.

In 2021, there will be 63 power battery investment and expansion projects in China (including fundraising projects), with a total investment (including planned investment) of more than 621.8 billion RMB, and the long-term planned new capacity has exceeded 2.5TWh.

From the perspective of specific capacity planning, leading companies have large-scale expansion plans, such as CALB, CATL, and Honeycomb Energy in TOP 10 lithium iron phosphate power battery manufacturers. The lithium battery capacity planning in 2025 has exceeded 500GWh.

Driven by the strength of battery companies, the demand for lithium battery equipment will further increase. It is expected that the market for machine vision equipment in the lithium battery field will maintain a high growth trend in the next few years.

It is worth mentioning that at present, lithium-ion battery machine vision manufacturers have also begun to create a full line of visual inspection solutions, such as Yihong Intelligent and APT.

Deep learning technology

The new demand for visual inspection of power batteries poses a huge challenge to traditional algorithms, and detection technology based on AI and deep learning will become one of the cores of competition for machine vision companies.

The underlying logic of the traditional detection algorithm is to find the difference in brightness and color in a designated area, and judge the size or shape of the foreign object accordingly. Obviously, the traditional algorithm cannot solve the detection pain points in the welding process, and cannot meet the detection needs in this field.

At present, deep learning technology has been gradually applied to the application of lithium-ion machine vision inspection. Compared with traditional algorithms, AI vision inspection systems can automatically learn and extract defect features to improve the detection rate. Deep learning has the advantages of high accuracy, strong robustness, and avoiding tedious feature design.

3D vision

3D vision is rapidly infiltrating in the field of power batteries, becoming a supplement to 2D vision, and the localization process is expected to speed up.

Battery companies are increasingly demanding machine vision requirements such as detection accuracy, speed, image transmission and defect analysis.

3D vision has ushered in a large-scale introduction period in the field of power batteries by adding depth information of physical space on the basis of traditional 2D vision.

It is expected that the field of 3D vision will continue to be highly popular in the next few years. The competition among global manufacturers will be more direct, and the localization process is expected to accelerate.