Drone team manipulation using hand gestures for object transportation

• Main Authors: Bayeta, Reginald Geoffrey Lausa, IV, Megino, Kyle Jomar Casabuena, Parco, Angelo Jose Teodorico Diaz, Vicente, Anjelo Louise Gerardo
• Format: text
• Language: English
• Published: Animo Repository 2022
• Subjects: Drone aircraft—Control systems; Gesture recognition (Computer science); Computer Engineering
• Online Access: https://animorepository.dlsu.edu.ph/etdb_ece/21; https://animorepository.dlsu.edu.ph/cgi/viewcontent.cgi?article=1019&context=etdb_ece
• Institution: De La Salle University

This paper presents the design and development of a Drone Team system that can be manipulated and controlled through hand gestures for object transportation. This system uses the Leap Motion Controller and Leap Motion SDK to read and measure hand data that can be interpreted and translated into gestures and commands. Additionally, the system was made considering the use of Crazyflie 2.0 nano-quadcopters in the drone team. This system is made and developed using MATLAB and Simulink alongside Robotics Operating System (ROS) and Gazebo. In coordination with Gazebo, ROS allows the drone team and the payload to be visually simulated while keeping track of location data necessary for control and data collection. MATLAB and Simulink are used to implement the various controllers on a Crazyflie 2.0 nano-quadcopter. Simulink would accept the data from ROS and process it based on the hand gesture command transmitted from the Leap Motion Controller. The Simulink model will then develop new desired destination data that will be transmitted to ROS with the change reflected in Gazebo. To test the system, the Leap Motion Controller will have its response time tested, and the system will be used to fly the drone team along three specified routes to test the ability of the system to control the drone team and payload. The tests show that the system can accurately and precisely control the drone team using hand gestures recorded from the Leap Motion Controller. Future studies are recommended to implement this system using real hardware outside simulations.