BIM-Based Trajectory Planning for Unmanned Aerial Vehicle-Enabled Box Girder Bridge Inspection

Abstract

Inspection is essential for bridge maintenance and has been supplemented by the use of unmanned aerial vehicle (UAV) photogrammetry. However, a review of the literature reveals that existing approaches require the intervention of a human operator to select waypoints in digital twin environments. Thus, existing studies are limited by either manual or semi-automatic control restrictions on UAV navigation settings, which is the main motivation for our work. This research developed a building information modelling (BIM)-based trajectory planning approach to enable fully autonomous UAV navigation for box girder bridge inspection, where the UAV follows a predetermined sequence to traverse the environmental space of a box girder bridge. The approach reflects the geometry properties of box girders as inspection characteristics and is designed with algorithms to determine a mathematical relationship between the box girders’ coordinates and inspection sequences. Field testing of the approach shows that it has satisfactory performance in terms of (1) navigation efficiency, (2) reasonableness of the sequence determined for trajectory feature points, (3) trajectory closeness and deviation, and (4) trajectory smoothness. Its satisfactory performance supports the practicability of fully autonomous UAV-enabled bridge inspection.