This thesis focuses on the study of the actuation properties of Aerial Vehicles (AV)s, and their ensuing feasible wrench sets and useful abilities. The the field of AVs, and the corresponding applications, expanded in the last decade, where many new AV designs emerged beyond the classical quadrotor design. These groups of designs employ a range of techniques to expand the platform’s abilities, such as actively tilting propellers while in flight to change the platform’s thrust direction, tilting platform’s propellers to achieve full actuation, or even optimizing the propeller’s placement and orientation to achieve omnidirectional flight. With the vast literature encompassing different designs, it was inevitable that each group of designs follows a specialized nomenclature and design framework. While this approach helped the advancement of these designs, it renders the comparison of the capabilities of different designs a challenging task. Moreover, while each design is demonstrated capable of the tasks it was built for, it would be interesting to define a set of basic abilities that could provide a clear idea of a platform’s possible applications. Unlike fixed wing platforms, AVs are used for their ability to hover in place, i.e., stabilize their position about a desired one over a period of time. From this configuration a platform should have the ability to then move around by following a desired trajectory. While these two abilities have been thoroughly discussed in the literature in an implicit and explicit way, we believe that the conditions for a platform to hover have been discussed either theoretically, or derived for specific platforms. However, a general numerical framework that allows the analysis of this ability was never introduced. Similarly for a platform’s ability to fly in an omnidirectional way: while the desired behavior of an omnidirectional platform is intuitively understood, and the conditions to achieve this ability have been discussed in the literature, these conditions usually ignore the platform’s actuation limits, and require a case by case analysis of the platform’s ability. From another perspective, AVs are usually equipped with an array of sensors that allow these vehicles to sense their environment and estimate their state. These sensors are the source of the intelligence associated with these platforms, and which allows them to navigate autonomously. One of these sensors, that has become crucial for the modern AV, is the Inertial Measurement Unit (IMU). The IMU provides acceleration, angular velocity and possibly magnetometer measurements. These measurements are crucial for the state estimation of the platform when flying indoor or outdoor. In addition to state estimation, new control schemes emerged in the last few years that attempt to benefit from the high frequency of these measurements to fly a quadrotor robustly. While these controllers have allowed quadrotors to achieve interesting performances, they still rely on high frequency measurements of the propeller speeds, and filter the IMU measurements with less than optimal filters. As such, and following the above introduction, this thesis attempts to provide a plethora of contributions both in the design and control perspective of multi-rotor aerial vehicles starting with modeling generic AVs, followed by numerical methods to analyze their basic abilities. These analyses are followed by contributions in the design of a novel omnidirectional prototype AV and by experimental analysis of different AV trajectory tracking. Finally, a novel method to filter IMU measurements and employ them in robust control of AVs. Each of these contributions is coupled with additional minor contributions, left for the curious reader to find. Finally, each of the theoretical contributions presented in this thesis is coupled with an extensive experimental campaign that demonstrates the stated hypotheses.