In civil engineering, monitoring structures is crucial for assessing potential damage. Challenges in integrating physical sensors with numerical simulations arise from the assumptions of 'perfect' conditions in simulations, which complicate damage predictions. This issue is significant in structures with unpredictable damage patterns, such as load-bearing glass façades, where the risk of sudden pane failure limits broader applications, despite the proven load capacity. The proposed solution is a 'soft sensor' or hybrid digital twin that merges physical sensor data with virtual models, and is significantly enhanced by machine learning to provide more accurate structural condition predictions. This approach enhances monitoring efficiency and accuracy, improves the reliability and safety of such structures, and promotes sustainable and efficient construction practices.     «

In civil engineering, monitoring structures is crucial for assessing potential damage. Challenges in integrating physical sensors with numerical simulations arise from the assumptions of 'perfect' conditions in simulations, which complicate damage predictions. This issue is significant in structures with unpredictable damage patterns, such as load-bearing glass façades, where the risk of sudden pane failure limits broader applications, despite the proven load capacity. The proposed solution is a...     »