Robotic mobile sensing for robust modal identification across a population of bridges: Uncertainty analysis, algorithm development, hardware realization, and field validation

Population-Based Structural Health Monitoring (PBSHM) has recently emerged as a promising paradigm to enhance monitoring capabilities across populations of structures. A central requirement for PBSHM is the collection of data from multiple population members. Conventional fixed sensing strategies, whether permanently installed or temporarily deployed, are impractical for this purpose, as they require significant time, labour, and cost. This study introduces a novel robotic mobile sensing framework designed to overcome these challenges. The framework develops a customized portable accelerometer and an intelligent wheeled robot carrying multiple sensors to conduct vibration-based measurements on bridge structures. The collected data enable modal identification, a cornerstone task in PBSHM. To address uncertainties inherent to mobile sensing, we conduct a theoretical uncertainty analysis and develop a robust automated frequency domain decomposition algorithm tailored for mobile data. The proposed framework, which encompasses sensing hardware, uncertainty analysis, and a modal identification algorithm, is validated through field deployment on ten simply supported bridge spans, representative of a bridge population. Using only two sensors, we successfully extract multiple modal frequencies and mode shapes for each span, while quantifying uncertainties in the results. Comparisons with finite element analyses and population-level assessment further confirm the effectiveness of the framework, highlighting its scalability, cost efficiency, and suitability for practical PBSHM implementation. • Uncertainty propagation framework for robotic mobile sensing in SHM. • Robust modal identification with statistical uncertainty estimates from mobile data. • GNSS-synchronized robotic sensing prototype for scalable vibration-based SHM. • First population-scale field validation of robotic mobile sensing on bridges.