Collecting dynamic bridge measurements with smartphone and GNSS technologies
Abstract
Last decade have seen a rapid increase in use of GNSS Navigation technology in the monitoring of flexible structures and bridges due to good reliability and low maintenance requirements. Its main drawback, however is the cost of the deployment, as dual frequency receivers with multipath resistant geodetic antennas are required to combat harsh environment. An alternative option is vision-based deformation monitoring, which deploys feature recognition, detection and tracking algorithms to collect information related to structural response, for example, movements of structural features such as bolts in steel bridges. This approach can offer lower cost deployment, especially if smart phone hardware can be used. This reduces safety risk to the bridge inspectors, allowing either for temporary or permanent installations of off-the bridge comparing with a GNSS antenna that has to be installed on a bridge. Installations of cameras could reduce potential risks of working at height and over busy roads and traffic disruption that are associated with the installation of GNSS sensors. Previous authors' research demonstrated that good and comparable results could be obtained using mobile phone cameras.
This study assess the performance of smartphone technologies in vision-based deformation monitoring using data collected on the Wilford Bridge, which crosses the River Trent and is both a footbridge and aqueduct bridge. The smartphone is placed on a tripod approximately 40 m away from the mid-span of the bridge. The frame size and collection rate of the videos are set to 3840×2160 pixels and 30 frames per second, respectively.
Visual-based data will be compared with data from two geodetic type GNSS receivers, monitoring the deflection of the two sides of the mid-span of bridge and collected at 10Hz frequency. Data is collected while group of pedestrians excited the bridge by walking, marching and jumping, causing semi-static and/or dynamic displacements of the bridge, and rotation of the deck, of different amplitude and frequencies. The analysis includes spectral analysis and band-pass filtering of the time-series, allowing to identify low- and high-frequency components expressing the semi-static and dynamic displacements. It will also identify the modal frequency obtained using smartphone and GNSS technologies, comparing it with results from authors' earlier work.
This study assess the performance of smartphone technologies in vision-based deformation monitoring using data collected on the Wilford Bridge, which crosses the River Trent and is both a footbridge and aqueduct bridge. The smartphone is placed on a tripod approximately 40 m away from the mid-span of the bridge. The frame size and collection rate of the videos are set to 3840×2160 pixels and 30 frames per second, respectively.
Visual-based data will be compared with data from two geodetic type GNSS receivers, monitoring the deflection of the two sides of the mid-span of bridge and collected at 10Hz frequency. Data is collected while group of pedestrians excited the bridge by walking, marching and jumping, causing semi-static and/or dynamic displacements of the bridge, and rotation of the deck, of different amplitude and frequencies. The analysis includes spectral analysis and band-pass filtering of the time-series, allowing to identify low- and high-frequency components expressing the semi-static and dynamic displacements. It will also identify the modal frequency obtained using smartphone and GNSS technologies, comparing it with results from authors' earlier work.
