The challenge

When processing high-value materials, the customer needed a system capable of detecting and mapping defects on their rolls, with several key expectations:

• Detect 100% of material defects and generate a defect map per roll
• Enable optimised nesting on the downstream machine to avoid defective areas
• Measure the exact dimensions of each roll and centralise data on the server

Overall, the challenge was to provide a reliable and accurate tool to prevent defective parts and achieve significant material savings.

The CNM response

CNM Process developed a defect mapping machine based on a camera inspection system and an embedded AI analysis model, with three key technical strengths:

• AI processing is embedded directly within the camera module, with no image transfer latency
• Analysis is carried out on raw data without JPG compression, ensuring maximum accuracy
• Detection of defects down to 0.2mm, with QR code and barcode reading for full traceability

Overall, the solution allows downstream machines to nest parts intelligently by avoiding defective areas, reducing waste and improving material yield.

The manufacturing process

The machine processes the roll through unwinding, inspection and data generation, with several key strengths:

• A 100% detection rate, with no processing delay thanks to the embedded AI
• Unwinding with PE liner removal for improved inspection quality
• Full roll mapping and automatic generation of the defect map

The advantages of our machine

The solution offers a 100% detection rate thanks to AI embedded directly within the camera module, with no latency or image compression. The defect map generated for each roll allows the downstream machine to nest parts while avoiding defective areas, thereby reducing material waste. QR code and barcode reading ensures full traceability and centralisation of data on the server.