Crowd-induced vibrations in stadiums pose challenges to structural integrity, human comfort, and long-term reliability. This paper presents a methodology that integrates ISO-weighted filtering, modal analysis, and statistical distribution modeling to evaluate vibration variability across three seating zones during a football game. Vertical vibrations are collected in the band, lower student, and upper student sections, which were processed to extract the vibration dose value (VDV). Modal peaks at 2.34 and 4.69 Hz were consistent in sections, with an additional 13.18 Hz mode observed in the band zone. Statistical fitting of Normal, Lognormal, and Weibull distributions to the VDV data yielded exceedance probabilities of 10.9%, 20.5 %, and 29.8 % for the band, lower student, and upper student sections, respectively. These results underscore the notable differences in comfort limits variations under identical game conditions. Future work recommendations are extended to apply the approach to multiple games, investigate causal links between variability and crowd-related factors, and quantify the contribution of each section to general discomfort to inform targeted design and maintenance strategies. The predictive capability of the framework offers a valuable tool for stadium monitoring and design optimization.

Bridge infrastructures lack affordable network-wide monitoring, as fixed S HM is typically limited to landmark spans. Drive-by monitoring with mobile robots offers scalable coverage using vehicles as moving sensor platforms. This paper experimentally evaluates a frequency domain subtraction algorithm on vibration data collected by an unmanned ground vehicle (UGV), validated against a portable SHM system on the UCF footbridge and the Alafaya Trail Bridge. The paired on-bridge and adjacent-road vibrations are transformed by FFT, scaled by energy and subtracted to isolate the dynamic response of the bridge and reveal modal peaks. On the footbridge, the UGV consistently recovered four of eight modal frequencies while on the stiffer Alafaya bridge, it identified the fundamental mode within 0.35% of the reference. Though portable SHM yields higher fidelity, it lacks network f easibility. UGV-based monitoring achieves a practical balance of accuracy and consistency. Future work will address uncertainty quantification, advanced modal frequency extraction algorithm and mode-shape extraction.