Hand Gesture-based Control of Autonomous UAV for Mobile Disinfectant Dispenser

Abstract

Autonomous unmanned aerial vehicles (UAVs) have witnessed a rapid increase in their utilization in various applications and will continue to do so in the coming decades. These UAVs, also known as drones, are designed to either assist humans or perform tasks that involve people. Drones of today have grown to be faster and less expensive by integrating several technologies, supported by hybrid algorithms, and perform various tedious, challenging, filthy and hazardous tasks. The deployment of machine learning and other AI-based algorithms enhances drones’ autonomous and vision capabilities. Today, part of an effort to curtail the spread of COVID-19, this research has designed, developed and built a mobile disinfectant dispenser based on autonomous quadrotor UAV. It is a “flying dispenser”, able to detect a person’s hand gestures from afar, based on machine learning (ML), to fly and maneuver towards the person and finally spray disinfectant on his/her hand. In order to identify various hand motions for maneuvering, this research studies and improves the ML algorithms and carries out various experiments to improve the drones’ response time and maneuvering performance, for the final objective of taking precautions to protect humans from Covid-19.