Answers are grounded in this paper and cite their sources.
Demonstrates effective spectral recovery in structural health monitoring using consumer wearables, highlighting the potential of crowdsensing.
The growing deterioration of civil infrastructure necessitates scalable and cost-effective Structural Health Monitoring (SHM) solutions. While crowdsensing using pedestrian-carried consumer devices (e.g., smartphones and smartwatches) offers a highly accessible alternative to permanent sensor networks, the acquired data is severely corrupted by human-structure interaction. Pedestrian kinematics from activities like walking, jogging, and bicycling, combined with biomechanical transmission paths, introduce complex spectral distortion that masks the structural modes, hindering reliable Operational Modal Analysis (OMA). To address this critical gap, this paper introduces a Physics-Guided, Mobility-Conditioned Fourier Neural Operator (PG-MC-FNO) designed to invert mobility-induced kinematic noise. By integrating mobility state conditioning and a physics-guided loss function, the deep learning architecture learns a mapping from distorted crowdsensing singular value spectra back to benchmark spectra. The proposed network was experimentally validated on an operational pedestrian bridge using pocketed smartphones and wrist-worn smartwatches. Results demonstrate that the PG-MC-FNO successfully filters out biomechanical interference, recovering bridge modal frequencies and achieving a spectral correlation score exceeding 0.99 against stationary benchmarks. Component ablation and baseline comparisons confirm that the physics-guided optimization, mobility conditioning, and FNO architecture are all essential for reliable spectral recovery. Ultimately, this methodology advances the viability of everyday consumer wearables for continuous and scalable crowdsensing-based SHM.
No takes yet. Share an insight, caveat, or question.
Luleci et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: