Title: Artificial Intelligence Across Disciplines: Research, Innovation, and Intelligent Solutions
Editors: Dr. Subita Bhagat, Dr. A. Balamurugan, Dr. P. Krishna Kumar, and Mrs. S. Nandhini Devi
ISBN: 978-81-69857-83-3
Chapter: 15
DOI: https://doi.org/10.59646/815/15
Author: Dr. Vijesh Joe C
Abstract
Modern civil infrastructure systems require continuous, automated diagnostic capabilities to identify structural degradation, mitigate catastrophic asset failures, and optimize life-cycle maintenance investments. This chapter explores the design and deployment of an autonomous, deep learning-driven structural health monitoring framework for intelligent infrastructure, focusing specifically on long-span cable-stayed bridges under variable environmental loading and dynamic vehicular excitation. By combining spatial-temporal graph neural networks with physics-informed recurrent autoencoders and variational anomaly detectors, the proposed methodology unifies vibration telemetry, ambient thermal variations, and acoustic emission feeds. Evaluated over eighteen months of empirical field data across 1,200 continuous IoT monitoring nodes, the model demonstrates high sensitivity to sub-surface micro-fissures and prestress cable slackening. The framework achieves a 99.1% true positive anomaly detection rate while reducing false alarms by 84.6% relative to classical baseline methods, providing an edge-computable paradigm for self-evaluating, resilient civil infrastructure.
Keywords: Intelligent Infrastructure, Structural Health Monitoring, Spatio-Temporal Graph Neural Networks, Physics-Informed Deep Learning, Anomaly Detection, Edge Computing.