Physics-Informed Machine Learning and Multi-Sensor Fusion for Structural Health Monitoring: A Bridge Case Study
Results show that multi-year monitoring data can be reduced into compact fatigue-relevant features while preserving traceability to raw measurements, and a supervisory agentic layer coordinates data-quality checks, multi-sensor consistency review, and confidence-tagged substitution, creating an auditable workflow for e...