PSI - Issue 84

Available online at www.sciencedirect.com

ScienceDirect

Procedia Structural Integrity 84 (2026) 73–80

© 2026 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the Conference Keywords: Existing bridges; Artificial neural networks; seismic risk assessment, GIS-based risk mapping, retrofit prioritization Abstract Bridges are key components of transportation networks, and their failure can have serious consequences for public safety, the economy, and mobility. As infrastructure becomes more exposed to both natural events and human-related risks, assessing these risks effectively is increasingly important. Traditional bridge management systems mainly rely on visual inspections, which are often slow and costly when applied on a large scale. One of the main challenges is moving from basic bridge data to more advanced risk analysis, which usually requires detailed and expensive surveys. Recent studies have shown that machine learning methods, especially Artificial Neural Networks (ANNs), can help in early estimation of bridge conditions. Building on a previously developed ANN-based framework for predicting structural degradation and traffic-related risks, this study expands the method to include earthquake risk. Seismic hazard information from maps, combined with bridge characteristics, is used to create a predictive model for quick and broad assessment of seismic risk. The method is tested on a sample of 95 highway bridges in central Italy. The results are shown using GIS-based maps, which help visualize risk distribution and support emergency planning and risk mitigation. This approach is adaptable and scalable, making it a useful tool for public authorities. It helps prioritize inspections and safety actions, supporting better, data-informed decisions for managing transport infrastructure. III Fabre Conference: Existing Bridges, Viaducts, and Tunnels: Research, Innovation, and Applications Regional-Scale Seismic Risk Mapping of Existing Highway Bridges Using Artificial Neural Networks Lorenzo Principi a *, Michele Morici a , Valeria Leggieri a , Andrea Dall’Asta b a School of Architecture and Design, University of Camerino, Viale della Rimembranza 3, 63100 Ascoli Piceno, Italy b School of Science and Technology, University of Camerino, Via Gentile III da Varano 7, 62032 Camerino, Italy

* Corresponding author. E-mail address: lorenzo.principi@unicam.it

2452-3216 © 2026 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the Conference 10.1016/j.prostr.2026.06.010