Multi-Sensor System for Pipe Inspection using an Autonomous Hybrid Aerial Robot

Autor: Marco A. Montes-Grova, Francisco J. Pérez-Grau, Antidio Viguria
Rok vydání: 2022
Zdroj: 2022 International Conference on Unmanned Aircraft Systems (ICUAS).
DOI: 10.1109/icuas54217.2022.9836225
Popis: The use of aerial robots for the inspection of Oil&Gas production plants has considerably evolved in recent years. This article presents an aerial hybrid robotic system with the ability to autonomously land on the pipe to be inspected, so it can deploy a satellite crawler robot for accurate inspections. In order to ensure safe autonomous navigation in unknown environments, the aircraft is equipped with a redundant localization system. A sensor fusion of multiple pose estimation sources is performed to estimate the aircraft pose in a robust and efficient manner. Regarding the detection of the pipe to land on, its position is estimated by using the point cloud of an on-board camera. This paper presents both the localization and the pipe detection modules. Finally, successful experimental results carried out in an indoor environment are shown, being all computations performed on board the aerial platform. A video of the results including a fully autonomous mission can be accessed at https://youtu.be/N3ZGVuDy1qA.
Databáze: OpenAIRE