Enabling Digital Twin for Bridge Deformation Analysis of Flash Flood Impacts: The Concept and Implementation Using DSLR Camera

A sustainable maintenance of bridges is a very challenging task. Accurate information on the bridges' structure integrity could inform their lifespans. Meanwhile Digital Twins have emerged as transformative tools in monitoring and analyzing structures by creating virtual replicas of physical bridges to predict structural behaviors under various conditions. The integration of digital twin technology with bridge structures has revolutionized the field of structural maintenance. However, this technology has not been suited yet to accommodate the usage of DSLR cameras to monitor deformation which has been a day-to-day practice in photogrammetric community. This paper explores the concepts and implementation of performing bridge deformation monitoring using a DSLR camera by employing a novel statistical analysis. A case study was conducted by photographing a bridge concrete pillar in two different epochs. A convergent imaging network was applied to each photographing campaign to shoot around 33 retro target points attached to the pillar's surface. Ten photos on each epoch are processed using the least squares bundle adjustment to compute camera parameters and points of coordinates. This type of imaging network produces an orthogonal structure of the resulted covariance matrices of the 3D coordinates to simplify the statistical displacement analysis. The orthogonal and homogenous covariance matrix structures simplify the congruency and localization test procedures. Also, a sub millimeter precision of the coordinate measurements is effortlessly attainable. It is revealed that there are insignificant structure movements occurring, and it has little implication of the bridge's internal structures. The study demonstrates the effectiveness of using DSLR cameras to enable digital twin to perform self-assessment mechanism to any deformation analysis on bridge structures.