Interactive Autonomous Aerial Robotic Systems - modeling, analysis, and control

Cutting-edge smart factories, within the framework of Industry 4.0 driven by the Internet of Things paradigm, are characterized by the automation of highly efficient manufacturing processes and product transportation, alongside the digitalization of these processes and their data flows. In this context, interactive autonomous robotics, as well as advancements in the capabilities of traditional fixed manipulators, represent the state-of-the-art technologies being researched. Autonomous robotics encompasses machines equipped with sensing, perception, communication, and decision-making abilities, as well as mobility in both known and unknown environments. Thanks to recent advancements, these machines are also increasingly capable of physical interaction with their surroundings for tasks like inspection, maintenance and cooperative transportation. Aerial robotics has rapidly become a key technology, enabling navigation in complex three-dimensional environments and performing tasks in hazardous or hard-to-reach areas. Two main approaches dominate the field: one involves enhancing traditional under-actuated platforms, like quadrotors, by adding manipulators, though this complicates control due to the coupled dynamics; the other focuses on designing fully actuated platforms, such as hexarotors with tilted propellers and simpler end-effectors, which shifts the complexity to the platform itself and is the preferred approach in current research. This Ph.D. thesis explores the design, control, and interaction capabilities of autonomous aerial robotic systems, with a focus on multi-rotor platforms. The research aims to enhance the actuation capabilities of these platforms, addressing the limitations of under-actuated systems in performing complex interaction tasks. Specifically, these tasks require both detection and subsequent control of the interaction, a concept referred to in this work as contact-aware interaction. The first part delves into the design and enhancement of actuation capabilities. It examines the limitations of under-actuated systems and proposes novel solutions for improving the maneuverability and control of multi-rotor platforms, emphasizing the need for fully actuated designs that are capable of executing complex interaction tasks. The second part focuses on navigation and control in industrial environments. It explores techniques for indoor localization and advanced control strategies tailored to fully actuated platforms. This section highlights the importance of precise control and reliable navigation, especially in dynamic and constrained scenarios where traditional approaches may be insufficient. The final part is dedicated to contact-aware interaction, specifically addressing the challenges of physical interaction between multi-rotor platforms and their environment. This section introduces innovative approaches to force estimation and control, providing alternatives to traditional sensor-based methods. Together, these three parts contribute to advancing the state of the art in aerial robotics, offering practical solutions to key challenges in design, navigation, and interaction.