UAV path checking in an aircraft inspection system

Abstract

Robots are swiftly developing from manufacturing unit workhouses to robot companions. They are used to perform tasks such as welding, assembly and handling raw materials [1]. Although robots are capable of understanding and interpreting situation, however, it has to go through the initial stage of localisation so to familiarise the environment. A mobile robot with zero knowledge of the surrounding environment is incapable to move independently. They will tend to get lost easily and this would result in making the robot wandering aimlessly. In addition, when defining a mission, obstacle avoidance is the crucial element of the whole system. Mobile robots should have the capability to avoid obstacles at all cost as it will be used in aircraft hangar environment. In the event that the mobile robot is unable to avoid any obstacles such as aircraft wires or human beings, it will result in high probability of robot making mistakes. The objective of this project is to integrate the mobile robot and drone by connecting them together with a long wire to help to inspect defects of the aircraft. Before robot making the first move, it should have the features for mapping the environment. Various methods have been presented over the years for mapping the environment and most early works relied on some method use such as Pose Tracking, Monte Carlo Localisation as well as H-SLAM [2]. Undeniably, the new state-of-the-art method would be Particle Filtering which have become very popular class of algorithms to solve estimation problems [3]. In addition, the drone will be flying around to take pictures of the defects parts and it may come in close contact with the aircraft and wires as well as the attached wire between them may tangle onto obstacle objects. Thus, it should equip with different state of the art technology such as infrared cameras, GPS and laser. The proposed method in this final report is to use obstacle avoidance algorithm which is implemented based on mathematics equations. The algorithm will be based on the given point cloud data file which is acquire from the localisation method to check if the path is clear from obstacles. The obstacle avoidance algorithm was tested based on the coordinate points from the point cloud data file. Inside the algorithm, it will compute the distance between the obstacle point and the intersection point where the obstacle line equation cuts through the line of equation that containing initial and final coordinate points. Computed distance less than 0.5m will be indicated as an unsafe zone whilst distance more than 0.5m will be indicated as safety zone.