DEPLOYMENT OF AI-SUPPORTED STRUCTURAL HEALTH MONITORING SYSTEMS FOR IN-SERVICE BRIDGES USING IOT SENSOR NETWORKS

Abstract

This study investigates the deployment of artificial intelligence (AI)–supported structural health monitoring (SHM) systems for in-service bridges using Internet of Things (IoT) sensor networks, a rapidly advancing domain that merges cutting-edge sensing, communication, and data analytics to improve infrastructure safety and durability. Traditional bridge inspections and early wired SHM systems have faced persistent limitations, including manual data interpretation, latency in damage detection, and scalability barriers due to cost and complexity. Recent breakthroughs in wireless IoT technologies, multi-modal sensor fusion, and AI-based analytics now offer the potential for continuous, automated, and highly reliable monitoring of structural integrity across diverse bridge typologies and environmental conditions. To critically synthesize the state of knowledge, this systematic review analyzed 146 peer-reviewed papers published between 2000 and 2022 spanning civil engineering, computer science, and information systems. The review explored key dimensions: (a) the historical evolution and objectives of SHM in bridge engineering; (b) IoT sensor modalities and network architectures enabling large-scale monitoring; (c) AI techniques for damage classification, anomaly detection, and signal feature engineering; (d) system reliability and data integrity strategies including calibration, drift compensation, and cybersecurity; (e) deployment challenges and scalability considerations across steel and concrete bridges; (f) comparative field case studies and lessons from global smart infrastructure programs; and (g) emerging research directions such as digital twins, blockchain data provenance, climate-resilient AI, and hybrid human–AI decision systems. The findings indicate that AI-enhanced SHM significantly improves predictive damage detection, reduces false alarms, and supports timely maintenance decisions, especially when combined with multi-sensor IoT networks and robust data governance. However, challenges remain in standardization, model retraining under concept drift, cost–benefit justification for large-scale deployment, and regulatory acceptance of AI-informed safety decisions. The review also highlights knowledge gaps around extreme climate resilience, secure and scalable data management, and human oversight frameworks for trustworthy AI. By consolidating insights across 146 studies, this work provides an integrated roadmap for researchers, engineers, and policymakers aiming to advance the next generation of smart, resilient, and cost-effective bridge monitoring systems.