Buildings are exposed to exterior climate conditions, and as a result, degradation phenomena such as frost damage, algae growth, salts, fungi, and wood deterioration can occur over time.

Using hygrothermal simulations, it is possible to predict fairly accurately the temperatures and moisture contents occurring in structures, but there is a problem with degradation criteria: existing degradation models appear to have little reliability because they are not systematically tested against reality, at best only in stationary laboratory conditions.

The goal of this project is to use spectral analysis, clustering, and fingerprinting to create a spatial mapping of occurred damage in a limited number of case studies. By comparing those damages with hygrothermal simulations, existing degradation models can be improved, which in the future does make it possible to make absolute statements about degradation risks.