Fragility Curves are crucial for understanding the vulnerability of buildings and constructions to earthquake-induced damage. They graphically represent the probability of a structure reaching or exceeding various damage states as a function of seismic intensity, aiding in developing effective retrofitting and modification strategies for risk reduction, insurance planning, and/or policy decisions. Fragility curves enable expedited assessment, which can serve as a rapid pre-selection criterion for massive datasets of civil structures and infrastructures. In this sense, they can be used as a ‘pre-screening’ tool for asset selection; the ones deemed more at risk can then be passed to more focused, structure-specific, and detailed analyses, or even selected for continuous monitoring and/or seismic amelioration. This approach is especially valuable in large-scale urban and regional studies, where assessing the seismic vulnerability of numerous structures individually would be extremely time-consuming and resource-intensive.

This study presents an innovative approach to assessing seismic risk by utilizing the Generalized Extreme Value (GEV) distribution to empirically model these curves, rather than the more conventional lognormal approach. Building on the sets of fragility curves developed by Rosti et al. in previous independent studies, this research work compares the newly developed curves, obtained using Python-based tools and real data from the Da.D.O. (Database of Observed Damage) platform, to these benchmarks and evaluates similarities and differences from them. This study aims to provide a more accurate and practical framework for rapid seismic risk assessment and informed decision-making by combining empirical modeling with real observed damage data.