Launcher manufacturing is undergoing a rapid transformation, driven by the need for increased production robustness, cost reduction, and enhanced sustainability. In particular, the adoption of automation and digitalisation in production processes has become essential to remain competitive in today’s space industry. This thesis contributes to this transition by developing a data-driven anomaly detection system for robotic drilling and riveting, a critical step in the assembly of launcher structures. Drawing inspiration from the field of audio classification, in which transformer-based models have recently achieved state-of-the-art results, this research proposes a novel adaptation of the spectrogram transformer architecture to the domain of industrial time-series sensor data. The resulting deep learning-based approach enables automated detection of anomalous behaviour through sensor signals, offering a scalable solution to process monitoring.

Research in miniaturization of satellite technology has enabled the development of a new class of satellites- PocketQubes. The extremely small form factor greatly limits power generation capabilities, with body-mounted arrays expected to generate around 1W average orbital power for the 3P size. Deployable solar arrays present an ideal solution for small satellites, enabling more demanding payloads onboard. Numerous designs for CubeSats have already been developed commercially and in academia.
This research focuses on adapting solar array designs available in literature to 3P PocketQubes, addressing their unique size, power and deployment constraints imposed by the deployer. A 4-panel array design in a wing configuration was designed, facilitating a peak power of 19.7 W. This was supported by a burn-wire mechanism and a torsion spring hinge. Structural analysis was conducted using LS-DYNA and ANSYS Mechanical to simulate deployment impact and launch vibrations, respectively. The analysis evaluated the design’s response to dynamic launch loads and mechanical impacts during deployment. The final assembly weight was 204.8 g, contributing to a specific power of 96.2 W/kg, comparable to many COTS deployable solutions for CubeSats. The procurement cost was found to be €1480 (excluding solar cell assemblies), and can be greatly lowered with higher order quantities for formation flying/distributed missions with multiple PocketQubes.
Additionally, this research examined the impact of solar array deployment configurations on power generation and orbital lifetime. Power generation was assessed using a simplified Python model, later verified through AGI STK simulations. Results indicated that the 𝛽 angle significantly influences power output across different configurations. Moreover, for peak 𝛽 angles representing noon-midnight and dusk-dawn orbits, seasonal variations were studied. This yielded an approximate 30% reduction during summer solstice for dusk-dawn orbits, but no visible change for noon-midnight orbits. For velocity aligned PocketQubes, configurations with panels mounted on the 5 × 5 cm face at a 135° deployment angle were found to be optimal. In contrast, for PocketQubes with pointing capabilities and high peak power requirements, the previously designed wing configuration was recommended.
The study of orbital lifetimes was facilitated by ESA’s DRAMA software and CROC was used to identify the minimum, average (random tumbling scenario), and maximum cross sections, contributing to a wide range of expected orbital lifetimes. In the velocity-aligned case, configurations featuring panels attached long edges (parallel to the drag force) yielded lifetimes suitable for long Earth observation missions. The latter configurations were found to be suited for shorter, technology demonstration missions. The launch date significantly impacted the results of the study, especially for configurations with higher ballistic coefficients. The results from this study however are highly idealistic, as the PocketQube angle of attack is expected to vary, largely influencing the drag area experienced. Assumptions regarding the drag coefficient and the use of NRLMSISE-00 drag model within DRAMA also limit the accuracy of the results, as seen when comparing the simulation and real observations from Delfi-PQ.

Solar cells have been used on satellites since early after their invention, now present on virtually every satellite. To accommodate the power consumption of satellites, large panels have to be fitted to it. These are normally folded in a harmonica-like way, to fit within the launch vehicle fuselage, and deployed when in orbit. Each solar array features different parameters, defining its deployment behaviour. While limited research has been performed on the behaviour of individual arrays, a lack of research into a generalised response to parameter variations can be observed. The research of this Master's thesis focused on the characterisation of solar array deployments, identifying the influence of array parameters on the deployment, and the differences between test, simulation, and operation environments. Additionally, the test results were used in a correlation study together with the results of a simulation model for validation and investigation of on-Earth effects.

Maximum Power Point Tracking (MPPT) is commonly used to optimize the power output of solar arrays for LEO satellites. Traditional MPPT techniques such as Perturb & Observe often struggle when the satellite is rotating and the temperature and incident irradiance are rapidly changing. This scenario frequently occurs with miniaturized satellites, such as the recent Delfi-PQ satellite launched by TU Delft. One possible solution utilizes temperature measurements to perform MPPT, which is a more responsive architecture and better suited to rapidly changing environmental conditions. This project characterized the performance of a triple-junction solar cell at various temperatures and irradiance levels. The performance of a linear temperature-based MPPT formula was then estimated by comparing the experimentally collected solar cell data to formula predictions. This was an important first step in the possible implementation of temperature-based MPPT on future Delfi satellites.

The thermal analysis of small satellites (SmallSats) is often given a low priority compared to other sub-systems, due to the short timelines and small budgets associated with SmallSats. However, the satellite's temperature is crucial for preventing mission failures. In this thesis, a new, open-source, modular software for thermal analysis of SmallSats is created with the Python coding language, to simplify and speed up the thermal analysis process. This modularity is achieved by: ensuring that any thermal model can be connected to any other thermal model with a single connect function, allowing a hierarchical model structure that is easily mutable and re-usable; integrating a material database into the software, allowing frequent re-use of common materials, coatings, and contact connections; and including a dedicated sensitivity analysis tool in the software to allow the user to assess the sensitivity of the model's temperatures on various input parameters. To assess the quality of the software's computations, they were verified with the commercial ESATAN-TMS software, and validated with flight data from the FUNcube-1 CubeSat. The validation showed Root-Mean-Square Errors ranging from 1.4 degrees C to 7.1 degrees C for the outside panels of the satellite. With a sensitivity analysis performed on FUNcube-1, it was found that the solar absorptivity of the spacecraft demonstrates the greatest influence on simulation outputs: the largest risk may be its degradation throughout time, due to ultra-violet and atomic oxygen exposure. If the satellite would have had white paint, a year-long degradation could lead to a 17.1 degrees C temperature increase of the maximum achieved temperature in the satellite, depending on the type of paint used. Further development into this software should aim to validate the model even further with different satellites, which will be an important step towards making this tool a standard for SmallSat thermal analysis.

To enable greater access to space, this study aims to develop a cost-effective PocketQube Deployment System capable of deploying multiple PocketQubes from various launch vehicles. Mission cost, of which launch cost is a significant driver, can be reduced by minimizing deployment system mass. This can be achieved by structural analyses of single and multi pod configurations using finite element method. A metric called ‘deployer ratio’, defined as the ratio of payload and deployer mass, is introduced to compare the effectiveness of CubeSat and PocketQube deployers. From the perspective of deployer mass and mission cost, it was observed that it is more efficient to launch PocketQubes from a multi pod deployer rather than a single pod deployer. Although the estimated launch cost for PocketQubes is currently equivalent to commercial manufacturers, further cost reduction is possible by optimizing the mass of such a multi pod deployer.

Cost and lifetime requirements are two of the key drivers for the development of complex systems for modern space missions. As a consequence, system reliability is a factor of primary concern, requiring appropriate analyses. A closely related topic is the study of the effects induced by failures to which the system may be subjected, necessary to identify criticalities and weaknesses within the system. Mitigation strategies may then be derived and implemented. The aforementioned activities are laborious, intensive and a significant number of iterations are needed to achieve a system design that fully complies with the dependability requirements. This research aims to perform an FMEA and reliability analysis of satellite solar arrays by developing a flexible modeling software for the simulation of the performance of any PVA configuration, able to provide a reliability assessment. These activities require the implementation of two different models: an electrical model and a reliability model.
As far as the FMEA is concerned, two types of analysis may be performed to study the effects of the failure modes. A static analysis comprises the investigation of the variations experienced by the characteristic curves of the solar array due to failures, while a dynamic analysis studies the response of the Solar Array Regulator (SAR). The outcome depends on the type of analysis performed: for the former, the characteristic curves of the healthy and faulty system are plotted. Then, the performance variations corresponding to the three main operating points (open-circuit, short-circuit, maximum power) are calculated. This allows characterizing the effects of the failures of the solar array. Furthermore, a severity classification of the failure modes is retrieved. The response of the solar array regulator is instead investigated through the dynamic analysis, where a load is connected to the solar array. Based on the wave-forms obtained, the correct implementation of the SAR may be verified and the sizing of other elements of the Electrical Power Subsystem (EPS) may be derived.
The reliability analysis, besides providing an accurate estimation of the system reliability, is also used to calculate the probability of occurrence of failures. From this, a probability classification is determined, leading to a criticality assessment of the failure modes for the selected PVA configuration. This allows highlighting weaknesses in the system and acting accordingly with the implementation of mitigation measures. Furthermore, through the sensitivity analysis, recommendations for reliability improvement may be derived.
The classification of the effects of the failure modes provides relevant insights about the synthesis of a reliable system, drawing attention to those components whose failure induces the most concerning performance degradation. Furthermore, the outcome of the research is a robust starting point for the development and analysis of a full EPS.