Smart-device-based crowdsensing of urban high-rise buildings during Super Typhoon Ragasa

Owing to the high cost and complex deployment of conventional structural health monitoring systems, smart-device-based crowdsensing has attracted increasing attention for structural response monitoring of high-rise buildings. This study investigates its feasibility and reliability through field measurements of eight high-rise buildings in Shenzhen, Guangzhou, and Zhuhai during Super Typhoon Ragasa in 2025. Smartphones, tablets, and high-precision accelerometers (HPAs) were employed to measure structural responses, and timestamp-based resampling was applied to correct the non-uniform sampling of smart-device acceleration records. Comparisons with HPA measurements show that smart devices can effectively capture the dominant low-frequency structural responses. For the extracted fundamental sway-mode responses, the correlation coefficients between the two types of measurements generally exceed 0.75. In addition, the natural frequencies of the two fundamental translational modes identified from smart devices agree closely with those obtained from HPAs, with relative errors generally below 0.5%. Smartphone videos were further processed using the Kanade-Lucas-Tomasi optical flow method to extract low-frequency displacement responses, with frequency-identification errors within 0.7% compared with HPA results. The analysis of both short- and long-term variations in structural dynamic properties further highlights the potential of smart-device-based crowdsensing as a rapid, low-cost, and scalable approach for monitoring urban high-rise buildings under typhoon conditions.