Tracking bridge condition over time using recurrent UAV-based inspection

With bridge owners and managers tasked with making major maintenance/repair decisions with inadequate funding and resources, there is a need to provide quantifiable metrics on the overall health of a bridge that can help the decision-makers prioritize bridge maintenance/repair projects. One important metric indicating the health of a structure is the changing rate of the structure condition over time. The condition rating from human visual inspection in the current bridge inspection practice is not adequate for evaluating the changing rate, because it is subjective in nature and severity and extent of defects is not rigorously delineated on the rating scale. Recently, with advanced sensor and flight performance, as well as more affordable prices, unmanned aerial vehicles (UAVs) have become a popular tool in infrastructure inspection practice. The enhanced controllability with precise global-positioning-systems and inertial measurement units and increased safety with omni-directional collision avoidance sensors has enabled UAVs to navigate in the otherwise inaccessible key spaces for bridge inspection. In this study, a methodology is presented using UAVs to conduct recurrent inspections of a bridge through flight missions. With reliable flight control applications coming into the market, the exact flight trajectories and camera locations can be near-perfectly replicated during each flight. This provides a unique opportunity to capture comparable images for the same area over time. Then a defect detection and quantification algorithm is developed to quantify the progression of the defect over time using a set of images for a similar location. The feasibility and efficacy of the proposed method is tested in an experimental case study of a split cylinder test.