Vision-guided Quadrotor Perching on Imperfectly Cylindrical Structures
S.A. McGinley (TU Delft - Aerospace Engineering)
S. Hamaza – Mentor (TU Delft - Control & Simulation)
More Info
expand_more
Outdoor Perch Detection Demo: In this video, the perch detection algorithm is tested outside in a complex environment. First, the environment is shown in the form of a point cloud stream taken from a stereo camera, and then the same scene is shown again, but this time with the algorithm running in real-time. The algorithm isolates the cluster in each frame that it finds to be the most suitable perch and visualises a segment of the cylinder that it has fit to this perch structures. The single colour point groups that are visible at certain points are a visualisation of other clusters resulting from the conditional clustering segmentation process.
https://youtu.be/Vmw2R6O6Z4kPassive Gripper Test: This video includes clips from the gripper tests validating the Slapper drone gripper to allow for passive perching on branches of different geometries. Successful perches are shown for each of the different branches used during the testing. These tests showed that the performance of the gripper is dependent on both perch structure geometry and drop height.
https://youtu.be/Rle8jImwDVUDirect Approach Validation Test: This video shows whether a direct approach to a detected perch structure is feasible. This is visualised using a box marker that is coloured green if a direct approach is feasible and red if it is not. The video shows that the feasibility is determined in real time and updates with changes in the environment and disturbances to the drone.
https://youtu.be/hoQh4WtlFPYPerch Tracking with Decoy Perches: Here, the algorithm that uses subsequent perch detections to improve the drones position above a perch after the initial detection is demonstrated. This is shown to work for changes in the perch's position and orientation and despite the presence of several decoy perch structures. Note that there is a delay before the drone updates its position due to the need to pause the simulation environment to change the perch position. The flight controller does not reconnect immediately once simulation is resumed and this causes the short delay.
https://youtu.be/9_x5Jpg2do0Slapper Perching Flight: The entire perching flight test with the Slapper drone is shown here. The algorithm is visualised with a video overlay which shows perch detections in real time. This was the same visualisation that was used for the manual approach point validation. This experiment successfully demonstrated the capability of the Slapper drone to autonomously detect a suitable perch structure and then perch with no prior knowledge of the environment.
https://youtu.be/ZXeYG8QJnfoGithub repository
https://github.com/BioMorphic-Intelligence-Lab/slapper-droneOther than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
The design of aerial robots capable of perching poses significant challenges, from requiring pilots to master precise manoeuvres, to devising hardware and software capable of adapting to diverse perch structures and complex field environments. The Slapper drone presented in this paper tackles these challenges through three main innovations. First, a lightweight, vision-based system for autonomous perch detection using onboard flight hardware detects (imperfect) cylindrical objects found in both natural and artificial environments. Second, an onboard flight planning algorithm autonomously handles the detection, approach and perching flight phases, removing the need for a pilot. Third, a completely passive gripper utilises bistable shell structures to allow for perching on general long narrow features without any precise control inputs or power consumption. This design was successfully validated through both simulation and multiple indoor flights to result in reliable autonomous quadrotor perching in real-world environments.