Software development by Lisker Oy relies on a deep understanding of technology, familiarity with sawmill industry processes, and a solid software architecture. CTO Panu Toukola explains how actual sawmill conditions, long software life cycles, and machine vision make up the basis for intelligent measurement and optimisation solutions.
“If a machine vision system does not understand wood, it is useless on a saw line. The system needs to see precisely, deal with dust, and generate data that can actually be used.”
Panu is responsible for software suites, product development, and the direction of future technology together with his team at Lisker. Lisker’s measurement technology combines machine vision, software architecture, and deep awareness of the processes used in the sawmill industry. This is why Lisker systems work reliably even in the demanding conditions prevailing at sawmills.
Sawmill conditions put heavy requirements on systems: dust, woodchips, moisture, sap, chemicals, water, and oil cause constant stress to technology. Lisker systems are designed from the ground up to work in precisely such circumstances. It is not only about shielding individual components, but about making sure already in the design stage that the entire system is serviceable and usable. Being easy to clean and service ensures proper functioning 24 hours a day.
“The functionality of systems is measured on the production line, not in a laboratory.”
Managing dust and dirt is critical for camera-based systems in particular. Hands-on experience becomes apparent in the end result: the tougher the production environment, the more important it is to ensure full protection.
“It is easy to make systems work in the laboratory, but actual success is measured by how they work on the production line – where woodchips fly and timber is on the move,” says Panu.
Simple shielding is however not enough. The system must also understand what is being measured, why, and how – meaning the structure of the logs, the quality of the sawn timber, yield optimisation, and identifying defects in the wood.
There are plenty of machine vision specialists out there, but few who have a deep awareness of the processes in place in the sawmill industry. A machine vision system is always embedded in a larger system, meaning that successful measurement requires knowing sawmill machinery and the secrets of the mechanical processing of wood. Lisker has an exceptional amount of such expertise, which is why their systems work without a hitch on sawing lines.
The key is to understand saw line mechanics and different sawing techniques. And when it comes to grading sawn timber, knowing different grades, fault types, and various plank cutting techniques is important.
This is the kind of competence that is required both in the development of algorithms and in the training of customers.
“A system will only yield its full benefits when both the software and the user understand how wood behaves and what underlying challenges there are to it.”
If there is a mistake in distinguishing a knot from a crack, a good piece of timber can be rejected – or a faulty piece can be moved forward in the process. This has a direct impact on yield and quality.
Machine vision in the sawmill industry consists of much more than just installing cameras on the line.
The reliable functioning of the measurement system and its ability to generate reliable data requires controlling for lighting, filtering, and the used wavelengths – meaning the way the surface of the wood is scanned to get all the required data out of it. This requires extensive theoretical understanding of e.g. optics, photonics, and the material properties of wood as well as experience in the design of solutions suitable for sawmill conditions.
Panu explains that each wood type and individual saw line has its own challenges, which is why systems need to adapt to changes in lighting and algorithms. There is no one solution that fits all. Lisker systems have been produced in a way that makes it possible to install the same devices for very different applications, depending on the type of wood, the sawmill conditions, and the production process. This makes it possible to implement different imaging and measuring solutions, while at the same time keeping tested software core components standardised.
“Laser technology leaves no room for guessing.”
Using various light sources, filters, and lasers is key to getting all the required data out of each piece. The right kind of lighting and filtering for the camera brings out wavelengths that reveal for instance surface structures, defects, knots, and branches. This is a science of its own, combining theory and practice in a demanding way.
Laser technology improves the reliability of measurements, especially when it comes to scanning the shape of the piece. When alignment, the measurement method, and the algorithm all work together, the result is a precise and full picture.
“Although shapes can be scanned even without lasers by using image details and depth analysis, this technology eliminates all guesswork, which is why Lisker has preferred to use it especially in applications that require precision measurements.
Lisker Oy
Vähäkuja 3 / 17200 VÄÄKSY / FINLAND
+358 3 872 3400 / info@lisker.fi / www.lisker.fi
