A high-level neural controller for leader-follower flight is presented. State of the art range-based relative localization schemes that rely exclusively on onboard sensors present an additional challenge to the leader-follower control problem since they restrict the flight conditions that guarantee observability. This novel controller was developed over an evolutionary process in which the simulation environment resembled the real-life constraints a group of MAVs would encounter. During the learning stage, a group of three agents is used, where one acts as a leader and flies a random trajectory, and the other two act as followers guided by a candidate controller that dictates the desired velocity commands. In the end, when equipped with the best-evolved controller, the follower agents are able to showcase a successful following behaviour that also enhances the observability of the system, although no observability metric was included in evolution.

There is no doubt that the smallest spacecraft platforms (CubeSats, NanoSats and microsats) are revolutionizing the market. The possibilities of developing spacecraft through standard platforms, commercial of the shelf components and then reduced the costs of development has allowed new organizations to enter the space section. On top of that, traditional businesses have been benefited by these type of new spacecraft, as more technology demonstrations and riskier missions have been able to be developed through this technology. Nevertheless, these spacecraft systems cannot compete with their larger counterparts due to their size limitations. The extra space will allow for larger spacecraft to accommodate more complex and precise instruments and payloads. Therefore, an individual CubeSat will never be able to compete with a larger system. Nevertheless, in the last two decades, the concept of using the combined power of these simple and cheap systems to perform a task only possible with larger spacecraft has been proposed. One of the most interesting possibilities among the use of these distributed space systems is the use of swarming. Swarming is a technique which consists of the combination of a large number of simple elements to allow for the emergence of properties that allow the whole system to act as more than just its parts. The benefits of spacecraft swarming are undeniable, as inherently swarming brings robustness, flexibility, and scalability to the system, all three properties extremely valued in space missions.\\ For the moment spacecraft swarming has been mostly based on generating formation patterns with a large degree of autonomy. Still, only a handful of techniques have been explored. This work aims to further advance the knowledge of spacecraft swarming techniques. To do so, the implementation of a swarming algorithm never tested in space will be studied The advantage of this algorithm is double. First, as all swarming techniques, it presents a certain degree of artificial intelligence that will allow the swarm to cope with unexpected situations. Second, by design, this algorithm requires really little knowledge of the swarm environment. This translates in less need for information processing between spacecraft and within each spacecraft. Since the application is focused on already spacecraft with limited capabilities, this is certainly advantageous. To fully understand the work here presented, it is recommended that the reader possesses graduate-level knowledge on orbital dynamics and control, as well as some undergraduate notions of computer programming and artificial intelligence techniques. If the reader lacks knowledge on any of the mentioned areas, it is recommended to read the associated literature study referenced in the bibliography.

In this paper a range-based relative localization solution is proposed and demonstrated in practice. The approach is based on wireless range measurements between robots, along with the communication of their velocities, accelerations, yaw rates, and height. It can be implemented on many robotic platforms without the need for dedicated sensors. With respect to previous work, we remove the dependency on a common heading reference between robots. The main advantage of this is that it makes the relative localization approach independent of magnetometer readings, which are notoriously unreliable in an indoor environment. A theoretical observability analysis shows that it may also have two disadvantages: the motion of the robots must meet more stringent conditions and the relative localization method becomes more susceptible to noise on the range measurements. However, simulation results have shown that in the presence of significant magnetic disturbances that are common to indoor environments, removing the heading dependency is beneficial. We conclude the paper by implementing the heading-independent method on real Micro Aerial Vehicles (MAVs) and performing leader-follower flight in an indoor environment. Despite the observability analysis showing leader-follower flight to be an especially difficult task, we still manage to successfully fly for over 3 minutes with two fully autonomous followers using only on-board sensing.