What role do vision systems play in automated quality assurance?

Camera-based image processing is the key to adaptive automation today. The ability to recognize objects more variably and to react to different shapes, positions and numbers simplifies product handling and feeding to the actual inspection process. In addition, image-based evaluation with AI support is now much more versatile, faster and more effective. Vision systems therefore create stable inspection conditions, robust and reliable data and high throughput rates.

The trend in the industry is towards small, individual batch sizes. How can optical and other quality assurance systems be adapted to this?

Small batch sizes mean more frequent retooling of systems. The more flexibly and easily QA systems can be adapted to these changes, the faster production can continue. The use of camera technology and image processing makes it possible to continuously analyze and react to changes in the situation, which means that QA tasks can be performed with fewer rigid specifications for different products. In combination with machine learning algorithms, the systems can be further improved to recognize product deviations and distinguish defects even more specifically, as object characteristics can then be classified even with significantly more variance.

Quality or speed in image recognition - which is more important for the manufacturing industry at the moment?

Fast systems that work cost-effectively with high throughput are of course always welcome. However, when it comes to multi-variant production with small batch sizes, it is important to utilize the systems efficiently and without long downtimes, without compromising product quality.

Modern inference systems such as IDS NXT Ocean are ideal for this and, thanks to user-friendly software tools, require little expert knowledge of AI and image processing. They therefore take into account another important point in the manufacturing industry: setting up, converting and commissioning the inspection processes should be as simple and time-saving as possible and can be carried out by anyone.

You can also read about "Quality assurance" in the December issue - click here for the e-paper of INDUSTRIAL Production 12/21