Forschungsthemen

Automated High-Throughput Analysis of Imaging Data

Modern photovoltaic systems generate large volumes of electroluminescence, infrared thermography, and visual inspection images. Extracting meaningful reliability information from these datasets requires automated and scalable methods. We develop algorithms and data pipelines for high-throughput analysis of imaging data, enabling statistical evaluation of large module populations and early detection of failure mechanisms. By linking imaging signatures to performance data, we identify degradation pathways and improve predictive lifetime models.

Infrared-Based Polymer Analysis (NIRA)

Polymeric components such as backsheets and encapsulants often determine long-term module stability. Chemical changes in these materials can lead to cracking, delamination, moisture ingress, and electrical failure. We apply infrared-based polymer analysis to enable rapid, non-destructive identification and classification of materials in the field. This approach allows us to detect degradation-related chemical changes, correlate material composition with failure patterns, and assess reliability risks across large installed fleets.

Ultraviolet Fluorescence

Ultraviolet-induced fluorescence imaging is a sensitive method for detecting early-stage degradation in polymeric and encapsulation materials. Photo-oxidation and additive depletion can produce characteristic fluorescence signals before macroscopic damage becomes visible. We use UV fluorescence diagnostics to identify subtle material changes and integrate the results with automated data analysis and field performance evaluation. This multi-modal approach strengthens our ability to detect emerging failure modes and improve long-term stability predictions.