Contact person
Niklas Lorén
Projektledare
Contact NiklasTo develop sustainable products and processes, we need a deep understanding of the internal structure of materials. With advanced synchrotron- and lab-based X-ray tomography, we can examine materials in 3D down to the microscopic level—without destroying them. This also makes it possible to study how materials change under different conditions.
X-ray tomography generates enormous amounts of data, and to make full use of this information, smart data management is essential. This is where AI comes in. By using AI—particularly methods such as Convolutional Neural Networks (CNNs) and score-based generative models (SGMs)—we can quickly and accurately identify key features in the images, a process known as segmentation.
The project is developing and adapting new AI algorithms to enhance 3D and 4D image analysis, including real-time segmentation. A key focus is on reducing noise and artifacts in the images to produce as clear and reliable information as possible.
These algorithms will support the participating companies in their material development. In doing so, the project also lays a stronger foundation for the broader industry to leverage X-ray tomography in the development of tomorrow’s sustainable materials.
The project is a collaboration between Billerud, Lund University, Tetra Pak, and RISE, and is funded by VINNOVA as part of the Advanced Digitalization program.
AI-Tomo: Faster material analysis
Active
Coordinator, project leader and research provider
Three years
14500000 SEK
Billerud, Tetra Pak, Lund University, Billerud, Tetra Pak Packaging Solutions, Lund University
Vinnova avancerad och innovativ digitalisering
Torsten Sjögren Tuerdi Maimaitiyili Torben Nilsson Pingel Martin Simonsson Aleksis Pirinen