AI in the Sky: Diverse Approaches to Drone Swarm Command, Control, Connection and Communication

Abstract

This master’s thesis investigates alternative control methodologies for unmanned aerial vehicles (UAVs), focusing on integrating advanced human-machine interaction techniques. Traditional remote-controlled and sensor-based systems pose limitations, particularly for users with physical disabilities. This study explores the potential of virtual keyboards, hand gesture recognition, and eye movement tracking as feasible alternatives to provide more accessible, intuitive control options for the swarm of drone operation. The research methodology encompasses the design and testing of three interaction systems. Each system utilizes computer vision and machine learning technologies to translate human gestures or gazes into drone commands. The virtual keyboard allows users to input commands through eye interactions, hand gestures are captured and processed to control drone movements, and eye movements are mapped to specific flight commands. Findings indicate that while these methods offer significant improvements in user accessibility and control precision, they also present challenges. These include the need for precise timing in eye interaction, inaccuracies in gesture recognition due to insufficient training data, and the potential for bias in command interpretation from eye movement datasets. The thesis discusses these challenges and proposes potential improvements, emphasizing the need for balanced training datasets and adaptive learning systems. It also explores the broader implications of this research for cognitive science and smart city applications, highlighting how enhanced UAV control interfaces could contribute to more autonomous and efficient drone operations. This work contributes to the understanding and development more accessible swarm of UAV control systems that leverage human-machine interaction technologies. While not groundbreaking, these advancements offer meaningful insights into the potential for more inclusive and responsive drone technologies in various practical applications.