The majority of all batteries produced worldwide come from Asia. However, many industries, especially the automotive sector, would prefer European production facilities for batteries in view of the current supply chain problems.

The manufacturing process for lithium-ion batteries and batteries based on other chemical components is relatively simple in itself: many process steps consist of coating thin carrier films with materials that are sometimes rare and therefore expensive, and using these to produce the anode, cathode and separator films required for the batteries. These must then be layered in the correct sequence and packaged in housings. If errors occur during these processes, it is usually difficult to recycle the materials used, as there are still hardly any industrially established and efficient processes for this. This makes rejects very expensive and the overall process quickly becomes uneconomical.

In order to reduce the error rate and thus the number of rejects, it is necessary to monitor each individual process step during the production of batteries and thus detect any errors that occur early on in the process. Mitsubishi Electric has a special technology for testing coated and uncoated films: many of the process steps described can be monitored using line scan bars or line sensors.

Distortion-free image capture

Line scan bars capture the image data of flat continuous materials and web material line by line and with a distance of just a few millimetres between the sensor and the test object. Their optical design enables distortion-free image capture, which simplifies subsequent evaluation. The line scan bars are used, for example, in the quality inspection of paper, textiles, sheet steel and other flat objects.

In battery production, the devices can take on the two main tasks in quality and process control: Firstly, they enable reliable and fast inspection of films for defects, scratches, holes or discoloration, which can lead to reduced battery performance. On the other hand, they are suitable for checking the dimensional accuracy of coatings as well as distances and other geometric specifications. Thanks to their telecentric design, line scan bars also solve the problem that the films and coatings used are often highly reflective.

More than just quality control

The battery manufacturing process comprises many process steps in which line scan bars from Mitsubishi Electric can be used for process control and optimization in addition to quality assurance.

© Mitsubishi Electric Europe

Based on the company's factory automation components, Mitsubishi Electric can create an end-to-end solution that optimizes the effectiveness of production processes. The line scan bars are the key component that can do more than just quality inspection: In cooperation with systems for image analysis, the line sensors can communicate in a standardized way as OPC-UA clients with higher-level OPC-UA servers in the plants and in this way allow direct feedback of results to the plant. The architecture developed by Mitsubishi Electric based on the OPC-UA standard enables the settings of the automation components involved to be adjusted quickly and thus provides immediate feedback as soon as measured values at any point in the process do not correspond to the specifications. Thanks to an immediate response from the plant, settings can be optimized both on automation systems in the plant and on the line scan bars, thereby reducing expensive rejects.

New findings for process optimization

In addition to the direct feedback of recorded sensor data into the system, the systems with the designation 'Qi<+>Pro' (Quality Inspection and Process Optimization) allow a further step towards greater efficiency in battery production: All recorded data can be stored in a cloud and analyzed there in order to bundle cross-site results and use them for all of a battery manufacturer's plants, for example. The background to this option is that, according to Mitsubishi, there is still no official catalog of faults worldwide that defines the influence of individual faults and their consequences, for example on battery runtimes and other battery parameters.

The author: Peter Stiefenhöfer is the owner of PS Marcom Services in Olching.

© Peter Stiefenhöfer

With the help of the data collected from the processes via the Qi<+>Pro systems, the basis is created for evaluating the data obtained using artificial intelligence. This enables the predictive simulation of systems and the optimization of all necessary processes. By evaluating the data recorded with the Qi<+>Pro systems, it is possible to subsequently analyze the results of changes to certain process parameters in order to draw conclusions for future improvements.