Artificially Inteligent Route Building Unsupervised Drone
Abstract
In the rapidly evolving landscape of drone technology, its application in surveillance, security, and rescue operations has the potential to yield groundbreaking solutions that help societal advancement. AIRBUD (Artificially Intelligent Route Building Unsupervised Drone) presents a quadcopter design, incorporating a CubeOrangePlus flight controller, an onboard Jetson Orin Nano computer, and a ZedX mini camera that contributes to that advancement. The overarching objective of AIRBUD is to autonomously achieve control objectives, including identifying, tracking, and path predicting the trajectory of a ball in flight. Live image information captured by the camera is processed in real-time on the onboard computer, facilitating automatic decision-making. Signals are then transmitted from the onboard computer to the flight controller, allowing the quadcopter to execute predefined control objectives. Leveraging neural network object detection algorithms, the system will demonstrate its capability to navigate around and find objects such as balls and people in an area. The communication protocol between components utilizes ROS to ensure efficient data transfer. In finality, this system will be able to artificially identify and track a ball with accurate yaw rate while predicting its location of initial ground impact.
Artificially Inteligent Route Building Unsupervised Drone
In the rapidly evolving landscape of drone technology, its application in surveillance, security, and rescue operations has the potential to yield groundbreaking solutions that help societal advancement. AIRBUD (Artificially Intelligent Route Building Unsupervised Drone) presents a quadcopter design, incorporating a CubeOrangePlus flight controller, an onboard Jetson Orin Nano computer, and a ZedX mini camera that contributes to that advancement. The overarching objective of AIRBUD is to autonomously achieve control objectives, including identifying, tracking, and path predicting the trajectory of a ball in flight. Live image information captured by the camera is processed in real-time on the onboard computer, facilitating automatic decision-making. Signals are then transmitted from the onboard computer to the flight controller, allowing the quadcopter to execute predefined control objectives. Leveraging neural network object detection algorithms, the system will demonstrate its capability to navigate around and find objects such as balls and people in an area. The communication protocol between components utilizes ROS to ensure efficient data transfer. In finality, this system will be able to artificially identify and track a ball with accurate yaw rate while predicting its location of initial ground impact.