GRASP: Gripping Robotic Aerial System for Pickup

Abstract

This project aims to add to the field of autonomous systems by designing, implementing, and evaluating an aerial robotic system equipped with a gripper intended to autonomously perform complex retrieval and delivery tasks. The quadrotor UAV, integrated with a simplistic robotic gripper, operates under a set of algorithms that enable it to autonomously launch, navigate to specific targets, execute grasping maneuvers, and transport objects to predetermined locations. The project addresses a growing need for sophisticated UAV capabilities, targeting applications such as commercial package delivery, cinematography, archaeology, emergency response, and infrastructure maintenance. The project delves into the quadrotor's design features, focusing on autonomous control technologies and mechanical adaptations that allow for precise manipulation and handling of objects. The system's performance was assessed through numerous experimental trials, revealing successful navigation and transportation capabilities, though highlighting limitations in reliable object retrieval under varying operational conditions. The discussion also explores the implications of these findings in the context of real-world scenarios, suggesting enhancements in sensor integration, control algorithms, and mechanical design to enhance functionality and reliability. Furthermore, the thesis proposes avenues for future work, including the refinement of the UAV's gripping mechanism and the enhancement of its autonomous decision-making processes to better adapt to the dynamic conditions of real-world environments. By adding to the literature of autonomous UAVs, this project lays the groundwork for future innovations by introducing an approach for autonomous aerial manipulation technologies.