A physics-informed multi-agent framework for real-time seismic health monitoring and structural safety management

Abstract Traditional seismic monitoring in active regions often delivers delayed or fragmented information, compromising real‑time structural assessments. To overcome these limitations, an agentic AI‑driven Seismic Health Monitoring System (AgenAI‑SHM), is introduced. This self-managing system leverages a physics-informed, multi-agent architecture designed for adaptive, human-in-the-loop structural oversight. The framework orchestrates five specialized agents (Seismic, Structural, Material, Alert, and Visualization) using the Model Context Protocol (MCP) for unified data semantics and the Agent-to-Agent (A2A) protocol for decentralized decision-making. Continuous synchronization with Building Information Modeling (BIM) environments is maintained via the Industry Foundation Classes (IFC) protocol. Functionally, the Seismic Agent detects anomalies in data streams, while the Structural and Material Agents evaluate real-time dynamic responses. Upon identifying abnormal vibrations, the Alert Agent transmits warnings via a human-in-the-loop validation interface. To ensure physical consistency, the framework integrates a Physics-Informed Temporal Convolutional Network (PI-TCN) to execute real-time inference of ground motion and state transitions. Validation using replayed records from the 1994 Northridge earthquake on a benchmark 10-story reinforced-concrete moment-resisting frame demonstrated the feasibility of the proposed real-time inference and decision pipeline. The system achieved classification scores of 92.5%-95.6% across precision, recall, and F1-scores, accurately tracing stiffness degradation up to 65%. Ultimately, AgenAI-SHM establishes a fault-tolerant, communication-aware paradigm for real-time physical AI reasoning in seismic health monitoring.

Authors

Institutions

Publication Details

Journal
Smart Construction and Sustainable Cities
Published
2026-11-14
DOI
https://doi.org/10.1007/s44268-026-00104-x
Primary Topic
Structural Health Monitoring Techniques
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

A physics-informed multi-agent framework for real-time seismic health monitoring and structural safety management

Artem Zaitsev
Smart Construction and Sustainable Cities
Structural Health Monitoring Techniques
article

A physics-informed multi-agent framework for real-time seismic health monitoring and structural safety management

Artem Zaitsev
article en

Abstract

Abstract Traditional seismic monitoring in active regions often delivers delayed or fragmented information, compromising real‑time structural assessments. To overcome these limitations, an agentic AI‑driven Seismic Health Monitoring System (AgenAI‑SHM), is introduced. This self-managing system leverages a physics-informed, multi-agent architecture designed for adaptive, human-in-the-loop structural oversight. The framework orchestrates five specialized agents (Seismic, Structural, Material, Alert, and Visualization) using the Model Context Protocol (MCP) for unified data semantics and the Agent-to-Agent (A2A) protocol for decentralized decision-making. Continuous synchronization with Building Information Modeling (BIM) environments is maintained via the Industry Foundation Classes (IFC) protocol. Functionally, the Seismic Agent detects anomalies in data streams, while the Structural and Material Agents evaluate real-time dynamic responses. Upon identifying abnormal vibrations, the Alert Agent transmits warnings via a human-in-the-loop validation interface. To ensure physical consistency, the framework integrates a Physics-Informed Temporal Convolutional Network (PI-TCN) to execute real-time inference of ground motion and state transitions. Validation using replayed records from the 1994 Northridge earthquake on a benchmark 10-story reinforced-concrete moment-resisting frame demonstrated the feasibility of the proposed real-time inference and decision pipeline. The system achieved classification scores of 92.5%-95.6% across precision, recall, and F1-scores, accurately tracing stiffness degradation up to 65%. Ultimately, AgenAI-SHM establishes a fault-tolerant, communication-aware paradigm for real-time physical AI reasoning in seismic health monitoring.

Smart Construction and Sustainable CitiesVol. 4(1)
Lomonosov Moscow State University (RU)
Structural Health Monitoring Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon