![[Translate to English:] KI für die Rohwarenkontrolle: Michel Heeke im Interview](/fileadmin/_processed_/b/5/csm_IMG_7287_df2dccf570.jpg)
Michel, you are the internal project lead for the CRAI project. What exactly is it about?
CRAI stands for Centre of Research and Development of Trustworthy AI Applications for Mid-Sized Companies. The aim of the project is to establish artificial intelligence (AI) applications in the SME sector – through cross-sector collaboration.
Why is the focus on cross-sector collaboration?
AI applications can often be used in a wide variety of fields and sectors. They therefore do not need to be developed from scratch every time. This saves both time and costs, whilst also enabling innovations to be driven forward much more quickly.
What role is Meyer & Meyer playing in the project?
Together with the DFKI, we are developing a solution for visual anomaly detection for raw material inspection at our competence centre, WKS Textilveredlungs-GmbH. On 20 April 2026, we were on site at the company headquarters in Wilsum and connected the technology there. The AI is designed to support and relieve the workload of staff by automatically detecting even the smallest defects in the raw material.
How exactly does the technology now in use at WKS work?
So-called ‘good images’ are captured of every bale that is unrolled. These are images of the goods with absolutely no defects. The AI then compares the raw material as it passes by with these images and checks for deviations. If a deviation exceeds a predefined threshold, the machine stops automatically. The advantage of this approach is that, in principle, any raw material can be used and checked by the AI.
You mentioned use cases in other industries: what might the cross-sector application of the technology look like?
Visual anomaly detection may also be applicable in other areas of application. This demonstrates that the technology developed is not limited to a single sector. Rather, a wide variety of sectors can benefit from the same AI approaches and utilise them to meet their respective requirements.
The installation of the technology in April was an important milestone. What happens next?
Over the coming months, we will be testing visual anomaly detection extensively during live operations at WKS. This will allow us to assess whether the technology meets our expectations. Only then will a decision be made on whether it will be permanently deployed in raw material inspection.