What question did this study set out to answer?

This review aims to synthesize the recent advancements in artificial intelligence and machine learning for enhancing bridge resilience during extreme conditions.

March 10, 2026Open Access

Artificial intelligence for bridge resilience under extreme conditions: in-depth analysis and future priorities

Key Points

This review aims to synthesize the recent advancements in artificial intelligence and machine learning for enhancing bridge resilience during extreme conditions.
Analyzed 308 publications from Scopus covering efforts between 2021-2025.
Used bibliometric mapping to outline research growth and themes.
Conducted an in-depth analysis of 20 key studies pertaining to bridge resilience and AI.
55% annual growth in AI research related to bridge health monitoring.
Hybrid deep models like CNN and LSTM improve damage detection accuracy significantly.
Study emphasizes the importance of interdisciplinary collaborations in shaping AI applications for infrastructure.

Abstract

• 308 publications analyzed to map AI advancements in bridge resilience. • 55% annual growth in AI research boosts bridge health monitoring. • Hybrid deep models (CNN, LSTM, attention) enhance damage detection accuracy. • Study reveals global, interdisciplinary collaborations shaping infrastructure. This review synthesizes recent advances in artificial intelligence (AI) and machine learning (ML) for bridge resilience under extreme conditions. Drawing on 308 Scopus publications from 2021–2025, it combines bibliometric mapping with an in‑depth analysis of 20 key studies to track research growth, themes, and international collaborations. Advanced AI models—convolutional neural networks, long‑short‑term memory networks, temporal convolutional networks, and attention‑based hybrids—enhance real‑time structural health monitoring, damage detection, and seismic response prediction. Multimodal data fusion and transfer learning further improve predictive accuracy and support proactive maintenance. Overall, AI‑driven approaches complement inspections and finite‑element models by providing rapid, data‑driven decision support for the design, monitoring, and management of bridges under earthquakes, floods, fire, and strong winds. The review identifies three priorities: embedding AI within digital twins and decision‑support systems for network‑level bridge management; extending applications beyond seismic hazards to a unified multi‑hazard resilience framework; and explicitly addressing data quality, uncertainty, and explainability so AI tools can be trusted in safety‑critical settings. An AI‑integrated bridge digital‑twin framework is proposed to link multimodal sensing, hybrid/PINN‑enabled prediction, and risk‑informed outputs for real‑time monitoring, early warning, and maintenance planning. Future work should strengthen interdisciplinary collaboration and the use of standardized, open datasets to accelerate robust AI deployment.

Read Full Paperexternally

KI fragen

Bookmark

View Full Paper