Artificial intelligence brings detection of wood characteristics to a new level

  • text size


Phone: +352 540 416
Fax: +352 540 417

In the coming years, artificial intelligence will become a major competitive factor in the woodworking industry. Weinig Luxscan, a German-Luxembourgish specialist for wood scanning solutions, has recently installed the first scanner with this kind of technology in Textuur’s sawmill in Viljandi/Estonia.

Luxscan, the leading specialist in automated timber optimizing and a member of the Weinig Group since 2007, has been providing customized optimizing solutions for cross-cutting and ripping for 22 years. Recently, the company has installed a scanner in a finger-jointing line, which with the help of artificial intelligence ensures the best possible yield.

Customized solutions

The Estonian sawmill decided to invest in the Combiscan Sense C400 from Weinig Luxscan. Siim Jõgi, the Head of Development at Textuur, explains: "We chose a scanner from Weinig Luxscan because our entire new line consisted of Weinig machines and we believed it would be good to have a scanner from Weinig as well. We were right about it."

For Textuur a good detection of defects on pine and spruce was of great importance. The CombiScan Sense C400 provides accurate detection and feeds the best cutting solutions to two optimizing crosscut saws, which process up to 40 boards of 3.5 m per minute at a speed of about 220 m/min. The artificial intelligence supports the scanner’s optimization software and helps laser and colour cameras detect defects such as sound knots, black knots, wanes, cracks, holes, edge defects, resin pockets, and blue stains. In fact, artificial intelligence optimizes not only the cutting process but also the planing process. The latter is done by the Front End Scanner from Weinig Luxscan. This machine recognizes the annual rings and thereby guarantees the correct alignment of the boards before they enter the planing machine.

AI learns wood detection

Opticore AI is an intelligent image processing software. This revolutionary method uses “deep learning”, a class of neural networks from the field of artificial intelligence, to automatically analyse and identify wood characteristics. According to Luxscan, there has been nothing like that in the wood industry so far. This new method results in higher detection accuracy because the software "learns" to process images like a human brain. With Opticore AI the scanner is trained how to recognize and classify wood defects automatically according to predefined parameters. Detection results and their repeatability with changing wood characteristics are now significantly better and the set-up time is shorter.

Based on the exact board data obtained during image processing, OptiCore allows programming of multiple qualities and zones tailored to customer’s final product requirements. This means that almost all requirements can be taken into account and almost any type of end product can be manufactured.

Jõgi is particularly satisfied with the high output quality and the return on this investment. "We have noticed that with the scanners supported by AI we have an advantage over our competition. The AI scanners provide accurate defect detection and work flawlessly," says Jõgi. In addition, the installation went very smoothly. Everything worked fine from the beginning of the project. The scanner training provided by Luxscan was very informative and well prepared.

JP Hildebrand, Managing Director at Luxscan proudly adds: "We’ve been developing and researching in the field of artificial intelligence for three years. Now, our customers can benefit from this practical solution."


Post comment