Fatigue Crack Detection Using Unmanned Aerial Systems in Under-Bridge Inspection

The applications of Unmanned Aerial Systems (UAS) for bridge inspection, with emphasis on under-bridge inspection and fatigue crack detection, were studied in this report. The potential benefits and challenges of using UAS for bridge inspection were identified through an extensive literature survey. The feasibility of using UAS for fatigue crack detection was studied by determining the minimum lighting, camera distance and environmental requirements for three UAS. The DJI Mavic UAS performed better than the others in both indoor and outdoor GPS-denied inspections. An in-service bridge in Ashton, Idaho was inspected using this UAS to find fatigue cracks. No fatigue crack, known or new, was detected in the UAS images but marker lines around the known fatigue cracks (drawn by the inspectors in previous inspections), concrete defects, and steel rust were detected. Thermography showed promising results for fatigue crack detection in a lab setting, but it was not feasible for UAS applications since it had to be performed using active thermography techniques to obtain adequate results. Additionally, image processing algorithms for autonomous detection of both concrete and fatigue cracks were successfully developed. These algorithms, especially for fatigue crack detection, require more images to perform better, but were demonstrated as feasible to aid in a real-time inspection