Autonomous Surface Inspection of Airplanes with Unmanned Aerial Systems

Abstract

Inspection of commercial aircrafts is a time-consuming and cost-intensive manual task. This paper describes the process to use an unmanned aerial system (UAS) to capture pictures for a visual inspection of a known object's surface, in this case an airplane, for later evaluation. The focus of this paper is on the process chain using an UAS for a visual inspection of an airplane's surface. The presented procedure starts by gathering a point cloud of the object by performing a laser scan in advance or using a 3D model. Following this, inspection poses are determined based on surface normal vectors. These poses are used to compute a time efficient path with a genetic algorithm. Finally, the obtained results are validated by simulation based on real hardware and it is shown that an inspection of an Airbus A320 can be achieved in desired time.