Bridges are critical components of transportation infrastructure, and their safety and functionality are essential for economic development and public safety. Considering increasing structural failures due to extreme events, data-driven tools have become increasingly valuable for supporting risk assessment and maintenance prioritization. This study builds upon a previously developed framework based on Artificial Neural Networks (ANNs), designed to rapidly estimate structural degradation and risk levels associated with traffic loads on existing bridges, in accordance with the new Italian Guidelines for Risk Classification. The framework is extended to address seismic risk, enabling the development of an ANN capable of estimating the seismic risk of existing bridges using only a reduced set of easily available data. The extended framework is applied to a case study involving 95 bridges state highway network in the Marche region, Italy, predicting seismic risk for each structure. Finally, the results are visualized through GIS-based geospatial analyses to explore the spatial distribution of risk, identify clusters of high-risk bridges, and highlight territorial patterns that can inform strategic planning. The findings provide a practical and scalable tool to support inspection prioritization and the development of risk mitigation strategies within large-scale Bridge Management Systems, promoting more efficient and informed infrastructure management under seismic risk.
Principi et al. (Thu,) studied this question.