Dynamic Interactive Digital Monitoring Systems for Bridge Management

Bridges are essential components of transportation infrastructure, and their safe and efficient management increasingly depends on monitoring strategies that extend beyond periodic visual inspections toward continuous and data-informed assessment. Within this context, this thesis presents an integrated, interactive, and digital framework for bridge monitoring and management, in which measured structural responses are systematically transformed into calibrated and actionable engineering information. The work is centered on operational modal analysis as a practical basis for vibration-based bridge assessment under operational conditions, and it combines sensing, signal processing, structural modeling, and automated interpretation within a unified methodological framework. The thesis first establishes the conceptual and methodological basis for dynamic bridge monitoring by clarifying the role of operational modal analysis within the broader evolution of bridge structural health monitoring toward digital, IoT-enabled, and management-oriented systems. On this basis, a custom-designed, cost-effective, and power-efficient wireless monitoring architecture based on MEMS accelerometers is developed for bridge applications. The proposed system integrates wireless sensing nodes, synchronization and preprocessing procedures, cloud communication, and a user-oriented graphical interface, enabling deployable and interactive monitoring. Its performance is validated through laboratory investigations and field deployment on a real bridge, demonstrating that the system can provide reliable dynamic measurements for modal identification and displacement estimation under realistic operational conditions. The thesis subsequently addresses the transformation of measured dynamic information into a calibrated numerical representation of structural behavior. For this purpose, an automated finite element model updating strategy is proposed, combining discrete wavelet transform-based denoising, sensitivity-guided parameter selection, design of experiments, and an artificial neural network-based inverse model implemented within a graphical user interface. This framework enables efficient calibration of finite element models using experimentally identified modal properties, thereby establishing a direct link between measured structural response and a digital analytical representation suitable for interpretation and future decision-support applications. In the final stage, the thesis advances toward automated structural intelligence through two successive artificial intelligence-based structural health monitoring frameworks. The first integrates automated modal identification, health-state classification, and damage localization within a unified end-to-end workflow. The second extends this approach through a condition-aware framework that explicitly accounts for environmental and operational variability and incorporates an additional verification stage to enhance robustness and reduce false-positive detections. Experimental validation shows that these strategies can reliably distinguish normal variability from structural change and can localize damage with high accuracy. The work presented in this thesis establishes a coherent progression from deployable bridge monitoring hardware to calibrated digital representation and automated diagnostic reasoning. Its principal contribution lies in the development of an integrated framework for bridge management that supports the transition from fragmented monitoring practice toward intelligent, digital, and scalable infrastructure systems.