Alleima

Smart Vision Technology Improves Quality Control at Alleima

Together with Alleima, Prevas has developed a machine vision solution for automated quality control of tubes during production. The solution uses AI-based image analysis to provide more consistent and efficient inspection compared with traditional manual visual inspection.

When high-quality standards call for smarter inspection

Alleima in Sandviken is a global manufacturer of high-value products in advanced stainless steels and special alloys. The company supplies industries with exceptionally high demands on quality and performance, making quality control a critical part of its operations.

The project was initiated to help Alleima ensure the quality of tubes used in demanding applications, including the oil and gas industry.

The products are long, typically between 12 and 16 meters, with very high requirements for surface quality. Previously, inspection was carried out manually, with an operator visually examining the tubes.

“It's a task that is both time-consuming and monotonous,” says Niklas Brusén, Vision and Systems Engineer at Prevas. “The results of manual inspection can also vary depending on factors such as the operator's experience, focus and level of attention at different times.”

Image source: Alleima

Image source: Alleima

Imaging and AI-based analysis in real time

The solution works by moving the tube forward along the production line while simultaneously rotating it. A camera captures the entire surface of the tube – generating up to 1,000 images for a single tube.

The images are processed in real time by an AI network trained to identify specific types of defects that can occur during rolling.

Any defects identified are then addressed manually, for example through grinding. The system therefore supports the production process rather than providing a fully automated solution.

“This makes the operators' work easier,” says Niklas Brusén. “Instead of inspecting the entire surface of the tube, they can focus on the areas that the system flags as potentially defective. This saves time and allows their expertise to be used where it adds the most value.”

A pilot project with significant potential

The installation at Alleima is a pilot project in which Alleima and Prevas are exploring the possibilities offered by advanced machine vision technology. So far, the focus has been on verifying that AI and neural networks can identify defects in a real production environment.

There are several opportunities to develop the solution further. One possibility is to map the XY coordinates of defects, making it possible to pinpoint exactly where they are located on the product.

“The solution has been verified in a real production environment and clearly demonstrates the value of AI and machine vision,” says Milad Abdhagh, Account Manager at Prevas. “Deeper integration with Alleima's other systems could create even greater value, for example by providing enhanced opportunities for traceability and production data analysis.”

Our Expertis

Production Development & Industrial Automation

Explore

Curious about our insights?

See all our insights