In order to apply for any of the following theses or internships, the candidate must have no more than 3 exams left.
In March 2025, the Hera s/c will perform a Mars flyby on its way to the binary asteroid system Didymos, which will provide an opportunity to refine the ephemerides estimation for the two Martian moons Phobos and Deimos.
The candidate will assess the expected performances of the orbit determination process by simulating Earth-based radiometric measurements (e.g., range and wide-band ΔDOR) and optical navigation images collected by Hera before, during, and after the flyby, using the current mission scenario. Furthermore, the candidate will perform a sensitivity analysis to assess how the measurement concept of operations affects the final trajectory reconstruction.
Uploaded: 16 Oct 2023/RLM
Code: LM23HERAEE
Optical images collected by deep-space probes are often used to estimate the relative position of the spacecraft with respect to their small body targets.
This thesis aims to perform a detailed literature review of currently available image processing and navigation techniques for missions to small bodies, with a specific focus on the LiciaCube mission to the binary asteroid system Didymos.
The candidate will develop a complete pipeline for the most common image processing techniques, leading to the extraction of the target’s center of brightness and limb profile from a given input picture.
The generated observables will be included within JPL’s orbit determination software MONTE and analyzed as part of the LiciaCube flyby reconstruction.
Uploaded: 9 Feb 2024/RLM
Code: LM24ODSNSB
The relative orbit of Rosetta around Comet 67P was reconstructed using the operational solution derived by ESOC’s Flight Dynamics team as part of the navigation process. The operational solution contains several discontinuities caused by the lack of modeling for the comet’s non-gravitational motion, which limits the applicability of such a solution for the ephemeris reconstruction of 67P.
This thesis aims to perform an accurate reconstruction of the relative orbit of Rosetta for specific mission phases, using a combination of Earth-based radiometric measurements (Doppler and range data) and optical images collected by the onboard cameras.
The candidate will develop high-fidelity dynamical and observational models using JPL’s orbit determination software MONTE and process these measurements within an orbit determination filter. The reconstructed orbit will be validated through comparison with the latest solutions found in literature.
Uploaded: 9 Feb 2024/RLM
Code: LM24ROSCCG
Hera is a European Space Agency space mission, aimed at conducting an in-depth investigation of the Didymos binary asteroid system following the impact of the DART spacecraft (NASA) on Dimorphos. One of Hera's primary objectives is to accurately estimate the mass and mass distribution of both asteroids. This entails determining the gravity field of Didymos and Dimorphos with precision, offering valuable insights into their overall mass and internal distribution.
To assess the expected accuracy in the gravity field of the asteroids, our approach involved tracking and modelling Hera, Juventas and Milani spacecrafts in orbit around the asteroids. This thesis introduces a complementary method to gauge the sensitivity to the asteroids' gravity field, following the approach adopted by OSIRIS-REx at Bennu. This method entails tracking, using spacecraft-based images, and modeling pebble-sized particles that might have been ejected from Dimorphos' surface after the DART impact, establishing sustained orbits, or following natural ejection. The candidate will utilize Python to model these particles and perform orbit determination using the MONTE (NASA-JPL) software.
Moreover, the candidate will conduct parametric analyses to evaluate the sensitivity of the results to key parameters. This involves exploring variations in particle size and number, as well as different observation schedules.
Uploaded: 7 Mar 2024/EG
Code: LM24HERAGD
Exploring small celestial bodies presents unique challenges, primarily due to their highly irregular shapes, which complicate the accurate estimation of their gravitational fields. This thesis investigates the application of an algorithm based on Convolutional Neural Networks for 3D mesh reconstruction of small bodies using optical images captured by a spacecraft in orbit.
The approach involves adapting Google’s Mesh R-CNN, a tool designed to reconstruct 3D meshes of objects from 2D images taken from varying orientations. The model will be trained using synthetic images and 3D mesh models generated in Blender, followed by testing on real images to evaluate its performance and applicability.
In orbit around Mars since 2004, the European Space Agency’s Mars Express spacecraft has conducted radio science experiments to study the planet’s atmosphere, surface, lithosphere and interior structure.
During its mission, Mars Express has performed more than 70 radio science bistatic radar (BSR) observations. These experiments use the spacecraft’s high-gain antenna to transmit a right-circularly polarized unmodulated carrier toward the specular reflection point between Earth and Mars, which spans various longitudes and latitudes as the spacecraft moves along its trajectory. The resulting specular reflections, collected by 70-meter ground-based antennas, contain relevant information about roughness and dielectric properties of the Martian surface. A subset of these BSR experiments targeted Mars’ polar regions, which are known to host stable caps of water ice mixed with Martian dust.
The project will be focused on the analysis of selected S- and X-band bistatic observations conducted by the Mars Express mission on Mars, with particular emphasis on the planet’s polar regions. The candidate will process raw data recorded at the time of observation, extract information on roughness and near-surface dielectric constant, and ultimately develop meaningful geophysical interpretations of the results.
Topic: Bistatic Radar / Data Analysis
Tutor: Giancorrado Brighi
Uploaded: 03/12/2024
Optical navigation encompasses most if not all deep space missions, either as a navigation tool or as a remote sensing technique to assess the surface properties of bodies, and even their interior. From the ESA/JAXA mission BepiColombo to Mercury, to the recently launched ESA spacecraft Juice en route to Europa and other jovian moons, most missions involve the analysis of images of planetary surfaces to extract scientific data. Planning for unvisited planetary regions or bodies with unknown small features requires the simulation of accurate and data-driven topographic features to test optical navigation methods and image processing algorithms. Still, simulating planetary and small bodies surfaces is not trivial, as the interaction of photons with surface materials needs to be appropriately simulated, both at the microscopic and macroscopic scale.
In this context, this master thesis project involves using path-tracing engines to simulate materials and geometries, implementing data-driven procedural geometry generation, custom reflectivity models for different moons, planets, or bodies, and query of spacecraft and camera trajectory and attitude. The candidate will use path-tracing software like Blender and develop automation routines using its Python API. Data from the Planetary Data System (PDS) will be used to validate the simulations, and comparison with real spacecraft images will be performed to assess the effectiveness of rendering and simulation.
Topic: Remote Sensing / Numerical Computation
Tutor: Andrea Togni
Uploaded: 28/01/2025
Radio occultation is a powerful technique for studying the ionospheres of planets and moons, typically yielding key information such as electron density profiles. Existing analysis methods predominantly rely on the Abel inversion. This thesis proposal aims to develop a new methodology based on a least-squares approach, which offers greater flexibility if compared to traditional methods.
The least-squares approach enables the simultaneous fitting of the radio signal's phase data to a parameterized ionospheric model, allowing for the inclusion of additional constraints and prior knowledge about the system. Also, it allows for a rapid estimation of the uncertainties associated with the derived electron density profiles compared to time-consuming approaches such as Monte Carlo analyses.
The proposed methodology will be applied to the analysis of radio occultation data from NASA's Juno spacecraft, focusing on the ionospheres of the Galilean moons (Europa and Ganymede). By deriving high-fidelity electron density profiles, this work will contribute to a better understanding of the complex ionospheric environments of these moons, as well as their interactions with Jupiter's magnetosphere.
Topic: Radio Occultations / Numerical Computation
Tutor: Andrea Caruso
Uploaded: 25/02/2025