SpecialitySwiss Federal Laboratories for Material Science and Technology, CH

Probing material complexity with computer vision-from pixels to propertiesThe characterisation of emergence through self-organisation and its presence and role in biological systems and bio-inspired materials is still not well established. We tackle this question by using a data-centric approach with the goal of introducing machine learning as the key for quantifying the level of emergence and for understanding complex systems in general. We use wood and its hierarchical structure as an example of a complex system and we show that it is possible to predict its macroscopic mechanical properties by accessing the information stored in the emergent fibre patterns. This is done using computer vision techniques on a large sample of wood lamellae images. We demonstrate that the accuracy of such predictions is dependent on the spatial scale examined. This provides us with a method of first identifying the presence of emergence and the then quantifying it.