A free Web App for Rapid Flood-Fragility Estimation of Masonry Walls
27/09/2026
We are pleased to announce the official release of a new interactive web application dedicated to the rapid estimation and sensitivity exploration of flood fragility for masonry walls.
Developed within the framework of STRENGTH, this digital tool brings cutting-edge research directly to engineers, researchers, and practitioners.
Assessing the vulnerability of masonry structures against flood actions traditionally requires computationally expensive mechanics-based limit analysis models.
To overcome these computational bottlenecks, this web application employs pretrained machine learning surrogate models trained on an extensive database of simulated wall cases.
The tool enables rapid screening, sensitivity analysis, and preliminary interpretation of wall responses under out-of-plane flood loading in real time.
The platform provides an intuitive, user-friendly interface divided into two primary analytical modules:
• The Single case prediction section evaluates one wall configuration at a time. Users select the boundary condition, number of upper floors, wall length, wall height, flood depth, and representative wall case. The app then displays a current configuration summary and provides the predicted fragility response.
• The Sensitivity heatmap section allows users to compare many wall configurations at once. Users can select a sensitivity-analysis mode, such as boundary condition versus wall case index or wall height versus wall case index. The app then generates a heatmap showing how the selected output changes across the chosen parameter range. To streamline input definition, the tool also incorporates representative wall cases that visually correlate numerical parameters with standard architectural typologies.
The application is freely accessible online via Streamlit Cloud (the link is in the right-hand column).
For further details, please refer to the attached document Web app for rapid flood-fragility estimation of masonry walls (rigth column).
Because the app is hosted on Streamlit Cloud, it may occasionally go to sleep after a period of inactivity. If this happens, users can simply click “Yes, get this app back up!” and wait a short time for the app to restart.
The web app was developed within the research study:
Georgiadis, S., Nale, M., Faggioli, M., Nikolić, Ž., and Benvenuti, E.
Machine Learning Surrogates for Predicting and Interpreting Flood Fragility Parameters of Masonry Walls.
AI in Civil Engineering.
