Multi-output modal identification of landmark suspension bridges with distributed smartphone data: Golden Gate Bridge

Bridge infrastructure assets possess ultimate value for safe, resilient, and sustainable transportation networks. Monitoring of bridge structural characteristics is an essential process to minimize damage-associated risk but requires expensive sensor instrumentation, manpower, and expert intervention. Besides, certain bridges' vitality exceeds practical needs due to their landmark identity with symbolic value. In this study, an economical and consumer-grade-distributed sensor array is utilized to determine dynamic characteristics of the Golden Gate Bridge, the most prominent landmark suspension bridge in the United States. The bridge is instrumented with multiple smartphones throughout the main and the side spans to collect vibration data without obstructing pedestrian or vehicle traffic. The accelerometer data collected under clock distribution are processed to retrieve modal frequencies and mode shapes of the bridge. Asynchronous and sampling-deficient sensing approaches are adopted to extract the bridge modal characteristics despite the low vibration frequency and amplitude of the long-span suspension bridge combined with limited sensing and acquisition quality of the smartphones. The findings show significant correlation with high-fidelity reference instrumentations and present the largest-scale civil infrastructure monitoring example utilizing smartphone technology.