Integrazione Dinamica di Modelli d'Altitudine per Corpi Celesti in Simulazioni Spaziali

Virtual reality applications are a vital part of the aerospace industry. Given the various difficulties, great costs, and ambitious aims of space agencies and companies, the employment of real-time 3D virtual environments provides a number of advantages to astronauts, researchers, engineers, and more. Virtual spaces can be built to plan missions or verify their potential outcome; virtual environments can be used to train astronauts or maintenance operators by exploiting features that different Extended Reality (XR) devices can offer. Physical-aware simulations can be fed with scientific data coming from powerful instruments like spacecraft, rovers and orbiters; some examples include ESA's Gaia (a spacecraft currently mapping millions of stars from the Milky Way and beyond) or NASA's MRO (Mars Reconnaissance Orbiter, in charge of scanning the red planet and recording data of its surface among other tasks). The need for accuracy in these contexts is critical, especially in all those simulations where a detailed representation of the surface topology is required: environment simulations for mission planning; deployment and operations of surface rovers; virtual reconstructions of specific landscape features for morphological studies; and many more. These use cases need to process terrain data as precisely as possible, with correctly georeferenced features that compose the celestial body's entire surface. The current work describes the process of integrating new features into Astra Data Navigator (ADN), a 3D software developed by ALTEC using the Unity3D game engine that allows for real-time navigation of the Solar system and the data provided by the Gaia star catalog, along with an accurate simulation of various spacecraft launched by ESA. Specifically, the work focuses on the visualization of real, georeferenced Digital Elevation Maps (DEMs) of celestial bodies' surfaces like Earth, the Moon and Mars. These maps are usually provided by research institutions in the form of raw data of high precision, which is quite large in size and thus requires particular care in being handled, especially in real-time applications where performance and responsiveness are of the utmost importance. First, a general review of virtual reality applications in the aerospace industry is provided, accompanied by examples and solutions adopted through the years. The focus then shifts to use cases more specifically related to the current work: tools for real-time navigation of the universe, alongside the visualization and 3D representation of real surface elevation data. After this, ADN and its core features are introduced. The main topic is then presented: on one hand, the tools and technologies used during the development, including an explanation of the scientific data being used and how it was processed; on the other hand, an in-depth analysis of the process behind the implementation of a working pipeline for integrating surface elevation data into ADN, including requirements and desiderata, the obstacles faced and their respective solutions. Finally, the results are shown, giving conclusions and discussing future developments. Supplementary activities have been performed in parallel with the core activities and addressed as well: namely, the (re-)implementation of active stereoscopy for ADN inside ALTEC's CAVE system, and the evaluation of a generative Artificial Intelligence model employed for inpainting incomplete surface maps.