Development of a smart-device-based vibration-measurement system: Effectiveness examination and application cases to existing structure

Abstract

Summary After the 2011 Great East Japan Earthquake, long‐term vibration measurement using high‐density instruments is one of the most critical issues for structural‐health‐monitoring owing to increasing deterioration and threat of future large earthquakes. Because of the high initial and running costs of traditional monitoring systems, smart‐device‐based measurement system is considered as a simple and easy solution. In this paper, the effectiveness of in‐built sensor, data transfer via wireless local area network, data acquisition to a synchronize cloud server, and trigger function using shaking table tests were firstly examined. A measurement system including a group of sensors has been established successfully based on the “control center” from which the trigger command can be send to other sensors immediately as any sensor/sensors is/are triggered. Then, the system is applied to seismic‐response and environment‐vibration measurement at existing structures. Results show that the observable acceleration level of smart devices is more than 5 gal in the frequency range of 0.1 to 10 Hz. The possible sampling rate is 100 Hz. Though it is unstable, correction methods have been proposed. Continuous measurement and data transfer is possible without data loss. Dynamic properties extracted from smart‐device‐based system is very similar to those extracted from high‐quality‐sensor‐based system.

A literature review of next-generation smart sensing technology in structural health monitoring

“…The application was tested under a real earthquake and compared with the conventional damage assessment techniques with more than 85% accuracy. Shrestha et al introduced the possibility of using multiple smart devices for vibration measurement and compared their results using high‐quality sensors. The proposed concept seemed analogous to decentralized wireless sensors; wherein, the smart devices were used for vibration measurement in place of conventional wireless sensors.…”

Deep Learning-Based Real-Time Auto Classification of Smartphone Measured Bridge Vibration Data

“…The last layer is the classification layer, in which the trained classifier is integrated with the iOS smartphone operating system for real-time and in-device classification. In this study, the iOS application program developed for acceleration measurement by the authors [41] was further extended to incorporate the trained model with predefined classification labels, so that it can classify the vibration dataset measured on different bridges.…”

“…In this study, for realizing end-to-end processing from raw observation data to analysis result, a framework for real-time auto classification of smartphone recorded bridge vibration signals was developed following the steps as illustrated in Figure 16. The iOS application program developed for acceleration measurement [41] was further extended to incorporate the trained model with predefined classification labels, so that it can classify vibration dataset measured on any other bridges in real-time. The integration of powerful machine learning models into Apps on iOS devices is possible due to Apple's straight forward machine learning framework known as "Core ML" (Core ML Apple developers [45]).…”

Image processing–based real‐time displacement monitoring methods using smart devices

“…Particularly, because such vision‐based measuring systems handle extensive image processing, ensuring the consistency of the sampling time is important. However, as a consequence of smart devices being a multitasking system and heavy onboard image processing, the image frame output data are not sampled at exactly equal intervals, as shown in Figure .…”