Use of Unmanned Aerial Vehicle for Quantitative Infrastructure Evaluation

Abstract

Unmanned aerial vehicles (UAVs) allow remote imaging which can be useful in infrastructure condition evaluation. Furthermore, emerging noncontact sensing techniques such as digital imaging correlation (DIC) and other photogrammetric and visual approaches, including simultaneous localization and mapping (SLAM), can compute three-dimensional (3D) coordinates and perform deformation measurements as in the case of DIC/photogrammetry. A quantitative assessment of ways remote sensing in conjunction with UAVs could be implemented in practical applications is critically needed to leverage such capabilities in structural health monitoring (SHM). A comparative investigation of the remote sensing capabilities of a commercially availabl’e UAV, as well as both an optical metrology system known by the acronym TRITOP and the X-Box Kinect, is presented in this paper. The evidence provided demonstrates that red-green-blue cameras on UAVs could detect, from varying distances, cracks of sizes comparable to those currently sought in visual inspections. In addition, mechanical tests were performed on representative bridge structural components to attempt, for the first time to the writers’ best knowledge, deformation measurements using an aerial vehicle; displacements and corresponding accuracies were quantified in static and flying conditions. Finally, an outdoor feasibility test with the UAV was accomplished on a pedestrian bridge to test the marker identification algorithm.