AI-Powered computer vision reshaping processes at Mondi

In the paper industry, similar to numerous other industrial sectors, process automation is increasingly gaining importance. This development results from the necessity to adapt to changes in the labor market as well as to meet increased demands for quality and efficiency. The integration of software solutions in the field of Computer Vision (CV) plays a central role by serving as an essential tool for automation and support.

Research in CV has made significant progress in the last years especially through the democratization and enhancement of the underlying technologies. This development opens a plethora of applications along the chain of production from incoming goods to final quality checks. The broad tool kit of CV poses enormous potential for a wide range of applications in the industrial field from process control to workers’ safety.

Particularly, functionalities such as classification, object recognition, and image segmentation are often considered highly promising, but also less popular use cases like classic anomaly detection are increasingly coming into focus. The use of CV can take pressure off operators and support them in their tasks. Especially monotonous, repetitive tasks, that are therefore prone to human error can be supported very well with CV. With new technologies conventional tasks around product or process inspections are becoming easier to tune to specific problems, which leads to more precise results in problem detection, less errors and in return to higher tolerances for users.

Another crucial aspect lies in the flexible adaptability of CV algorithms to specific work processes as well as the utilization of existing infrastructure. This last point, in practice, allows for the adaptation and utilization of existing camera systems, often avoiding the need for new equipment acquisitions.