Wood prediction using Artificial Intelligence in Modelling (WAIMO)
|Funding from Vinnova||SEK 600 000|
|Project duration||September 2019 - March 2020|
|Venture||Strategic innovation programme for process industrial IT and automation – PiiA|
|Call||Digitization of industrial value chains|
Purpose and goal
Pulp manufacturing is affected by a large number of variables & dependencies and has for a long time been controlled using data. This data enables the possibility to describe these dependencies and thereby increase the processes efficiency, quality and optimize resource use. To achieve this, it is crucial to combine process knowledge and analytical models with new possibilities using AI. This feasibility study will run as a pilot project with the aim to predict the fiber wall thickness. A crucial property for the products of which the pulp is used for.
Expected results and effects
The project has the following goals: 1.Create a predictive model for the fibre wall thickness on the Iggesund softwood line. 2.Evaluate the potential for AI supported modelling of pulp properties. We believe that this in the future will give us a more stable and efficient production with a more optimized resource usage.
Planned approach and implementation
The project will be conducted on the Iggesund pulp mill and is a collaboration with a producer (Holmen), technical suppliers (Andritz & PulpEye) and research institutes (RISE Bioeconomy & RISE SICS). The purpose of the collaboration is to create a fuller understanding of the ingoing material, the process, measurement data and modelling. The feasibility study started in September 2019 and is expected to progress in 7 months. Besides creating a predictive model one will also evaluate the potential in this project concept.