Artificial Intelligence for Small Satellites Mission Autonomy

Space mission engineering has always been recognized as a very challenging and innovative branch of engineering: since the beginning of the space race, numerous milestones, key successes and failures, improvements, and connections with other engineering domains have been reached. Despite its relative young age, space engineering discipline has not gone through homogeneous times: alternation of leading nations, shifts in public and private interests, allocations of resources to different domains and goals are all examples of an intrinsic dynamism that characterized this discipline. The dynamism is even more striking in the last two decades, in which several factors contributed to the fervour of this period. Two of the most important ones were certainly the increased presence and push of the commercial and private sector and the overall intent of reducing the size of the spacecraft while maintaining comparable level of performances. A key example of the second driver is the introduction, in 1999, of a new category of space systems called CubeSats. Envisioned and designed to ease the access to space for universities, by standardizing the development of the spacecraft and by ensuring high probabilities of acceptance as piggyback customers in launches, the standard was quickly adopted not only by universities, but also by agencies and private companies. CubeSats turned out to be a disruptive innovation, and the space mission ecosystem was deeply changed by this. New mission concepts and architectures are being developed: CubeSats are now considered as secondary payloads of bigger missions, constellations are being deployed in Low Earth Orbit to perform observation missions to a performance level considered to be only achievable by traditional, fully-sized spacecraft. CubeSats, and more in general the small satellites technology, had to overcome important challenges in the last few years that were constraining and reducing the diffusion and adoption potential of smaller spacecraft for scientific and technology demonstration missions. Among these challenges were: the miniaturization of propulsion technologies, to enable concepts such as Rendezvous and Docking, or interplanetary missions; the improvement of telecommunication state of the art for small satellites, to enable the downlink to Earth of all the data acquired during the mission; and the miniaturization of scientific instruments, to be able to exploit CubeSats in more meaningful, scientific, ways. With the size reduction and with the consolidation of the technology, many aspects of a space mission are reduced in consequence: among these, costs, development and launch times can be cited. An important aspect that has not been demonstrated to scale accordingly is operations: even for small satellite missions, human operators and performant ground control centres are needed. In addition, with the possibility of having constellations or interplanetary distributed missions, a redesign of how operations are management is required, to cope with the innovation in space mission architectures. The present work has been carried out to address the issue of operations for small satellite missions. The thesis presents a research, carried out in several institutions (Politecnico di Torino, MIT, NASA JPL), aimed at improving the autonomy level of space missions, and in particular of small satellites. The key technology exploited in the research is Artificial Intelligence, a computer science branch that has gained extreme interest in research disciplines such as medicine, security, image recognition and language processing, and is currently making its way in space engineering as well. The thesis focuses on three topics, and three related applications have been developed and are here presented: autonomous operations by means of event detection algorithms, intelligent failure detection on small satellite actuator systems, and decision-making support thanks to intelligent tradespace exploration during the preliminary design of space missions. The Artificial Intelligent technologies explored are: Machine Learning, and in particular Neural Networks; Knowledge-based Systems, and in particular Fuzzy Logics; Evolutionary Algorithms, and in particular Genetic Algorithms. The thesis covers the domain (small satellites), the technology (Artificial Intelligence), the focus (mission autonomy) and presents three case studies, that demonstrate the feasibility of employing Artificial Intelligence to enhance how missions are currently operated and designed.