UAV inspection has become a trend in application ranging from search-and-rescue to maintenance. The later will be the main focus of this research proposal and the deployment of a UAV to be able to visit predefined goal locations around a building, a bridge, or a construction site and inspect for cracks, corrosion, or whether wear out has occurred around the inspected area. Anomaly detections will be logged and reported to an inspection manager for maintenance actions.
Autonomous local docking-navigation of a UAV has many applications throughout the scope of this vehicles outreach. SLAM (VSLAM/Lidar SLAM) poses a ‘honing-in’ docking method to safely navigate and avoid obstacles in a stable, controlled manner. Many docks (landing zones) are not situated on a fixed point in space, demanding a thorough computational docking system that quickly adapts to an ever-moving target. To achieve safe, stable and controlled docking and decent, the algorithms that induce such SLAM methodologies require quick computing times and thorough extremities for the ever-changing, demanding nature of the exercise. In pursuit of this success, the most opportunistic 3D-SLAM methodology (vision or laser) must be applied, taking into consideration effects of noise, range, frequency and accuracy.
Robot Operating System (ROS) offers a wide-range of UAV models that incorporate different hardware renditions available in the area of focus. It is the ideal tool to test and impose different systems onto a UAV model, trailing different strategies with alternate sensory capabilities. Therefore, ROS will be ideal in the use of the above autonomous scenarios for inspection and docking as a full circular (open-ended) solution, which will be cost effective and time saving for the construction industry.