The performance of Remaining Useful Life (RUL) prediction models is often limited by data scarcity, especially in safety-critical systems like aircraft engines where failure data is rare. To address this challenge, we propose the Super-SpaceTime GAN, a framework for generating synthetic condition monitoring (CM) data to enhance RUL predictions. The framework incorporates dual-conditioning on operating conditions (OCs) and RUL labels, an autoencoder-based latent space for denoising, and a supervised loss function to align synthetic data with real degradation trajectories. Evaluated on the CMAPSS FD004 dataset, the SuperSpaceTime GAN generates synthetic data that closely mimic real distributions, as verified using JensenShannon Distance, Principal Component Analysis, t-distributed Stochastic Neighbor Embedding, and a novel autoencoder-based health monitoring metric. The framework demonstrates improvement in prognostic performance in limited training data scenarios, with gains persisting even in data-rich settings. These findings highlight the potential of the Super-SpaceTime GAN to improve RUL predictions by addressing data scarcity, making it a valuable tool for Prognostics and Health Management (PHM).

Inverse design is a concept where one can design a structure via a property-first approach, where the properties of a system act as precursor information to guide the search for a viable design. This concept is classified as one of two: indirect or direct. While the former is no different than a forward-based approach using traditional optimization, a direct inverse design framework attempts to invert the process and directly map the desired properties to a suitable design candidate, which is unconventional. As such, the design process can be positively radicalized, with regards to the required computational resources when designing multiple structures.

In this work, the use of Bayesian machine learning, more specifically Bayesian optimization using Gaussian process regression, is employed to construct a direct inverse design framework using the design of mass-optimized fixed-wing aircraft ribs against buckling as the validation case study in 1, 6, and 10 dimensions. The results are compared to an indirect approach using the same algorithms and demonstrate that barring certain limitations, which are not inherent flaws of the direct approach, the framework designed not only demonstrates sound potential in inverting the forward map, but outperforms the indirect approach when designing multiple structures.

This research equally provides a stepping stone towards future research possibilities in the same field, all culminating in the improvement of the multi-disciplinary design process.

This paper presents an encoder-decoder-style convolutional neural network (CNN) for the purpose of improving monocular and stereo depth estimation (SDE) estimates, by combining them with the corresponding monocular estimates through a fusion network, assisted by prior information to provide context for the fusion. Video cameras are commonly used for depth perception in robotics, especially weight-sensitive applications, such as on Micro Aerial Vehicles (MAV). The two primary paradigms for vision-based depth perception are monocular and stereo depth or disparity estimation, each having their own strengths and weaknesses. These strengths and weaknesses seem to be complementary, and thus a fusion of the two may result in more accurate predictions. In this paper, we investigate this fusion by training a CNN that combines stereo and monocular depth or disparity estimates. The fusion network is agnostic to the choice of the input networks, providing great flexibility. It was found that such a fusion network, while increasing the computational complexity of the depth perception pipeline, indeed improves the accuracy of the estimates. The number of outlier predictions has been significantly decreased, while also limiting some fundamental limitations of both stereo and monocular methods, such as errors arising from occluded regions.

Military ground vehicles require protection against ballistic threats encountered in the battlefield. Traditionally used armour steel dramatically increases the weight of a vehicle, decreasing its mobility and hampering its capability of carrying payload. The development of new material and protection concepts, namely non-metallic armours, that function as the main structural element of the vehicle hull aims to optimise this relationship between protection and manoeuvrability, mainly through weight reduction. In this work, the ballistic impact response of multi-layered non-metallic targets under the impact of high-speed armour-piercing projectiles is studied by experimental and analytical methods. The residual velocity of the projectile, the failure modes and the energy absorption of different target configurations composed of aluminium oxide, E-glass or carbon fibre-reinforced polymer composite and aluminium plates are analysed. The results show that ceramic/composite target configurations containing aluminium interlayers have better anti-penetration capability against 12.7x99mm APM2 (hardened steel core) and 7.62x51mm AP8 (tungsten carbide core) projectiles, and E-glass outperforms carbon fibre in composite backings. An energy-based terminal ballistics model is built by combining two analytical models from literature and modifying them based on experimental observations as a cost-effective, simple modern approach to forecast the ballistic performance of ceramic-faced composite-backed targets. Model verification proves challenging due to incorrect formulation, lack of detail in the algorithm and insufficient input data provided by the authors. The ceramic plate submodel is verified, and projectile and ceramic damage mechanisms contribute the most to the projectile kinetic energy dissipation. There are significant disagreements between experimental and analytical results, mainly deriving from incorrect input values of projectile and ceramic material properties. The accuracy of the analytical results is assessed by quantifying and propagating the uncertainties in model input parameters through the present penetration model. An 8mm aluminium oxide/23.04mm E-glass composite target is predicted to have as high as a 100% chance of complete perforation by a 7.62x51mm AP8 projectile, having an areal density comparable to an 8mm aluminium oxide/6.35mm aluminium/11.52mm E-glass composite target which proved to have a 60% chance of defeating the same threat. This finding highlights the advantages of using predictive analytical modelling to assist armour design processes and make risk-informed decisions.

Drones have been an emerging trend in the last few years. They are used in multiple industries already, from photography and videography to racing. More use cases are now being conceived, such as using drones to deliver packages and food to people at home, using drones for inspections, or even using them as rescue searching vehicles in hostile environments. Even more possibilities open up once the drones bundle their forces to create swarms. The lifting capabilities of drones are still somewhat limited, but in a swarm they might be able to lift heavy payloads. This report covers the design of a concept of a payload carrying swarm, intended to lift cargo up to 500 kg to even the top of a tall building.

As the demand on new lightweight materials increases due to economic and environmental reasons, fiber reinforced plastics seem an interesting solution. In order to maximize their potential, ways have to be found to introduce damage tolerant designs. To this end, in the present exercise, the damage caused by cyclic loading is investigated.
Cyclic loading causes loss of stiffness and strength of composite structures. In the present work a progressive damage analysis model has been built. Damage is tracked by a macroscopic failure criterion. Failure occurring at levels below the macroscopic level are captured by a degradation theory which requires S-N curves as input. An attempt is made to use S-N curves derived from static tests only. Stress solutions are obtained using a finite element package. In order to bring the macroscopic damage, degradation and stress solution parts together, a cycle jumping algorithm has been developed which minimizes the amount of FEM stress solutions required and the required amount of function evaluations.
In its current form, the model is able to properly predict the trend of stiffness and strength behavior. Showing that a model like this has potential. For the proper predictions of mag- nitudes, additional failure modes should be modeled and the usage of S-N curves obtained from large specimens for applying on smaller geometries (elements) should be reconsidered.