Rizauddin Ramli

Afflilation: National University of Malaysia
Country: Malaysia
Congress: ICCS2021

Autonomous Navigation of Unmanned Aerial Vehicle by SLAM

In recent years, unmanned aerial vehicles(UAVs) have become more popular due to the rapid increase in affordability and their ability in aerial photography and decrease the costs needed for inspections on complex infrastructures. It is an emerging technology in Industrial Revolution 4.0 that have potential as disruptive innovation in a large number of industrial and civil applications. UAV’s popularity in industries has been increasing rapidly because of it’s ability to reach high-risk position which is difficult to access and hazardous for human. A huge potential market for this application is constituted by high-risk industrial infrastructure including oil and gas(O&G) industry, chemical plant, power station and shipbuilding. It is a crucial task to achieve accurate and precise navigation for UAV, in an outdoor environments with a lots of disturbance and obstacles. In this study, an autonomous navigation algorithms for a multi-rotor hexacopter UAV will be presented based on visual fiducial system and 3 dimensional Simultaneous Localization and Mapping (SLAM) in outdoor environments. The navigation algorithm of the autonomous UAV is divided into two stages. First, a 3D SLAM package is used create a 3D map of the environment based on the point cloud data captured using a commercial 360o laser scanning sensor attached to the UAV. The created map is called for localization of the UAV real-time control and used as the feedback of the proportional integral derivative(PID) control system of navigation. A path from origin to target is set as reference and the UAV follows the trajectory while the PID control system converges the state-state error. Next, after the UAV reached its desired position, an Intel® RealSenseTM depth camera will detect an April Tag attached to the target. In this second stage, the autonomous navigation will be based on orientation and distance between the camera and detected tag. The control system converges the state-state error based on the actual data from visual fiducial system. Finally, after the UAV reached its final target, a return home path planning based on the SLAM navigation is carried out.

Rizauddin Ramli will be speaking at International Congress on Control Systems 2021 which is scheduled to happen on 13th and 14th August 2021 at Hong Kong, HKSAR.