@phdthesis{,
    author = {Caroselli, Edoardo},
    title = {Autonomous LiDAR-free Navigation and AI-assisted Landing on Small Celestial Bodies},
    editor = {},
    booktitle = {},
    series = {},
    journal = {},
    address = {},
    publisher = {},
    edition = {},
    year = {2024},
    isbn = {},
    volume = {},
    number =  {},
    pages = {},
    url = {},
    doi = {},
    keywords = {Autonomous Navigatio Artificial Intelligence Hazard Detection And Avoidance Planetary Landing Asteroids Machine Learning},
    abstract = {Exploring small bodies in our solar system, such as asteroids and comets, has become a significant focus in planetary sciences due to their potential insights into the Solar System’s formation, resource utilization prospects, and the need to assess and mitigate planetary defense risks. In situ exploration and sample return missions to these small bodies provide valuable information about their physical and chemical properties, composition, and environmental conditions. Autonomy is crucial in advancing the boundaries of small solar system body (SSSB) missions, enabling spacecraft to perform tasks and make decisions independently. This thesis aims to contribute to the advancement of autonomy in SSSB missions, focusing on the precise landing of micro-spacecraft. The research investigates critical aspects of autonomous navigation, focusing on environment perception and decision-making. LiDAR-free vision-based navigation and AI-assisted landing techniques are explored for far-range and close-range navigation. The thesis presents the following main contributions: Firstly, an autonomous vision-based relative navigation system is designed, implemented, and tested. This system allows spacecraft to navigate in the proximity of the surface using relative measurements. It employs a novel monocular simultaneous localization and mapping (SLAM)-based filter assisted by altimeter measurements, enabling pinpoint landing at the target landing site. Secondly, an AI-assisted autonomous safe landing site selection technology is designed, implemented, and tested. This technology fuses image processing and machine learning methods, requiring minimal user input and incorporating landing requirements directly into the algorithm. Authentic mission images are used for validation. This research provides insights into the challenges and opportunities in achieving autonomy in SSSB missions. The thesis concludes by summarizing the main contributions, discussing limitations, and suggesting future research directions to further enhance spacecraft navigation capabilities in exploring small bodies in our solar system.},
    note = {},
    
    school = {Universität der Bundeswehr München},
    
}