Open Access

REVIEW

Artificial Intelligence in Earthquake Engineering: A Systematic Review from Machine Learning to Agentic AI

Jui-Sheng Chou^*, Dani Nugraha Limantono, Asmare Molla

Department of Civil and Construction Engineering, National Taiwan University of Science and Technology, Taipei, Taiwan

* Corresponding Author: Jui-Sheng Chou. Email:

(This article belongs to the Special Issue: Machine Learning Applications in Earthquake Engineering: Advances, Challenges, and Future Directions)

Computer Modeling in Engineering & Sciences 2026, 148(2), 4 https://doi.org/10.32604/cmes.2026.084591

Received 25 April 2026; Accepted 22 June 2026; Issue published 28 August 2026

Abstract

Artificial intelligence (AI) is increasingly transforming earthquake engineering by supporting prediction, assessment, monitoring, and decision support. However, existing studies remain fragmented because machine learning (ML), deep learning (DL), physics-informed AI, hybrid models, large language models (LLMs), multimodal AI, digital twins, and agentic AI are often examined separately. This study presents a systematic literature review of AI applications in earthquake engineering, synthesizing 130 studies published between 2016 and 2026. The reviewed applications include seismic hazard assessment, earthquake prediction, earthquake early warning, ground-motion modeling, structural response prediction, damage detection, bridge and building assessment, structural health monitoring, geotechnical earthquake engineering, post-earthquake reconnaissance, and decision support. This study develops an integrated taxonomy and evaluates each AI paradigm in terms of strengths, limitations, maturity, interpretability, scalability, robustness, and deployment readiness. The findings show that ML is effective for structured regression and classification tasks, DL is suitable for waveform, image, sensor, and time-series data, and physics-informed and hybrid AI improve physical consistency. LLMs, multimodal AI, digital twins, and agentic AI extend AI toward knowledge interpretation, interaction, workflow automation, digital-twin-enabled integration, and human-supervised decision support. Key challenges include data limitations, weak generalization, uncertainty, explainability, cybersecurity, ethical–legal concerns, and deployment readiness. Overall, this review outlines pathways toward trustworthy, resilient, and deployment-ready earthquake-engineering AI systems.

---

Keywords

---

Supplementary Material

Supplementary Material File

---