UAV-based Autonomous and Instantaneous Bridge Diagnosis

Abstract

Unmanned aerial vehicle (UAV) has been widely utilized in various engineering fields. This paper provides an overview of an UAV based bridge inspection project sponsored by Ministry of Land, Infrastructure and Transport in South Korea. This project aims to complement existing bridge inspection practices by integrating UAV, non-destructive testing sensing and big data analytics. The specific objectives of the project consist of (1) development of a UAV hardware platform capable of close-range flight near a bridge and flight-to-attach operation onto the bridge (40 minutes flight time with 4.5 kg of payload under 2 m/sec wind speed); (2) autonomous UAV operation using a vision based SLAM algorithm (under 50 mm location error) for localization and a line-of-sight path control algorithm under gust wind conditions (less than 4m/s); (3) hybrid imaging system combining infrared and vision-based cameras for visualization and quantification of various defects (micro-crack, spalling, bolt loosening and coating thickness variation); (4) automatic generation of 3D BIM model and overlapping with bridge inspection results; and (5) development of UAV-based bridge inspection guideline and regulations. Component level validation tests were performed on concrete and steel bridge structures, and testing of the entire integrated system is underway.