Neuro-symbolic Twin Networks for Autonomous Infrastructure Health Monitoring and Failure Forecasting Integrating Explainable Ai With Physics-guided Graph Neural Models for Real-time Decision-making in Smart Civil Structures
Year: 2025
Type: article
Abstract: The integration of Neuro-Symbolic Twin Networks with physics-guided Graph Neural Networks (GNNs) offers a transformative approach to real-time infrastructure health monitoring and failure forecasting in smart civil structures. This research investigates how hybrid AI models leveraging symbolic reasoning and physical constraints can enhance decision-making accuracy and transparency. The paper emphasizes explainable artificial intelligence (XAI) integration for interpretability, making these systems viable for safety-critical applications like bridges, buildings, and dams. By embedding domain knowledge into learning architectures, the proposed framework achieves higher forecasting accuracy, interpretable insights, and adaptive responses to structural anomalies.
Author: Researcher
Language: English
Subfield: Civil and Structural Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6950011734
FWCI:
Cites:
Cited by:
Related to: