Machine learning to boost quality at Rauma sawmill

​At the new Rauma sawmill, incoming logs are graded to assign them to the most appropriate purpose and enhance quality consistency in end products.
  • Articles
  • |
  • |
  • 2021, Rauma sawmill, Sawn timber

To optimise the sawing process to a precision that was previously unattainable, the mill will use measuring equipment and an information system to manage and learn from the measurement data.

The comprehensive range of X-ray and laser measuring devices are to be supplied by Finnos, a Finnish company specialising in measurement technology and software solutions for the forest industry. The devices will feature a control system that supports production planning. It will compile all the data generated during sawing, enabling the continuous improvement of process efficiency and optimisation of the process itself with the help of artificial intelligence.

The traceability of data within the sawmill plays a key role in the new system.

The control system grades logs into sawing batches and links them to the intended end products. The data accumulated during sawing indicates the kinds of boards obtained from a single log, and the part of the log from which the boards were sawn. The system also compares the data on sawing with the quality of the end product. This helps determine whether the goals set in grading were achieved in production.

Using a comprehensive dataset and machine learning, the system will learn to optimise grading and sawing.

“The goal is to improve the efficiency of sawing and to ensure that the quality of the sawn goods is as consistent as possible. The decisions made during log grading will also become more accurate,” says Jere Heikkinen, CEO of Finnos.

Unique system for the Rauma sawmill

The measurement and information system to be introduced at the Rauma sawmill is unique in scope and offers significant advantages. It will promote efficient and appropriate use of the wood and provide customers with consistent quality products suitable for the intended end use.

“Ours will be the only system in the world featuring technology as advanced as this. It is no exaggeration to say that the Rauma sawmill is the world’s most modern,” says Heikkinen.

Automated measurement and data collection contribute to the sawmill’s speed and large capacity. The Rauma sawmill receives approximately 7.5 million pine logs annually, and each of their individual properties can be taken into account in sawing.

Measurement points on the production line from grading to sawing

The new sawmill’s production line will feature several measurement points.

In log grading, a laser measures the log’s outer dimensions and an X-ray inspects it to determine all the features affecting the quality of the end product, such as knots, growth ring width and wood density. Calculations and analysis are carried out continuously while the log is measured.

09-Hero-Metsa-Fibre-sahatyomaa-Rauma-250821.jpg

Based on the measurement data, the log is assigned to the production of an end product that best suits its properties. The goal of grading is to maximise the economic value of logs and the yield of raw material.

Each log receives a unique digital ID, which enables the logs to be linked to the end product, as well as to the correct incoming batch. This provides the system with information about the felling site where the log originated.

After debarking, the log moves on to an X-ray measuring device that identifies the log based on its internal wood structure.

The sawing machine is also equipped with a measuring device that scans the surfaces processed by the saw and monitors the quality of sawing. The data generated during sawing is linked to the data of each individual log.

In this article

FINNOS

  • Manufactures measuring devices and software for the sawmill, pulp, energy and plywood industries.

  • Established in 2016

  • 30 employees in Finland

  • A turnover of EUR 3.6 million in 2020