More effective use of data supports both customer satisfaction and environmental responsibility.
AI uses measurement data from sawing
Metsä Fibre’s sawmill technological innovations, such as self-learning AI, support more cost-efficient production. At the Rauma sawmill, a virtual fingerprint is created for every log based on the log’s size, shape and inner quality using laser measurement and log X-ray technology. In the sawing process, a log is identified based on its virtual fingerprint, after which AI recommends the best possible sawing method. “Log sorting and measurement generate an enormous amount of data. AI allows us to combine data generated by several measuring instruments, compile it in one place and use it to optimise production and logs,” says Jarkko Vihervuori, Senior Development Manager at Metsä Fibre’s Rauma sawmill. AI makes it possible to select individual logs from the stream of raw material and sort them into different quality categories, enabling Metsä Fibre to produce targeted product batches. This also improves the sawmill’s utilisation ratio since the use of every log is optimised. “From the customer’s perspective, this is important because we can now more precisely allocate the desired logs to the specific end product,” says Vihervuori.
combinations. In spreadsheet software, a calculation on this scale would take several days. Metsä Fibre uses linear multivariable optimisation for calculations. Data on sawn timber supply, production and demand is entered into the tool, which then calculates the best way to use the raw material. Next, the data is entered into the allocation tool used by Metsä Fibre’s sawn timber sales organisation. This tool shows the availability of different sawn timber products and is updated automatically when a batch is sold. The introduction of optimisation and allocation tools reflects the fundamental change that is taking place in Metsä Fibre’s sawmill business. Metsä Fibre’s production is now based on real customer demand instead of historical data or statistics. Consequently, the company is able to better serve its key customers by responding more quickly to changes in customer demand and by improving production security. The digitalisation of production and demand planning also improves the efficiency of raw material use and reduces the carbon footprint. “When we understand customer demand better, we can focus on producing products that meet the actual customer demand. At the same time, we can ramp down the manufacture of products with lower demand or a longer supply chain and larger carbon footprint,” says Mäki. Striving for industrial efficiency allows Metsä Fibre to respond better to market demands and optimise production, thereby enhancing the company’s competitiveness. At the same time, more effective use of data supports both customer satisfaction and environmental responsibility.•
Algorithms support sawn timber production and sale
If the full production potential of log raw material was specified and organised in a spreadsheet, it would encompass tens of thousands of rows. Choosing the best possible method for production optimisation requires a complex calculation with hundreds of thousands of different
22
FIBRE
FIBRE
23
Powered by FlippingBook