Urban Air Mobility (UAM) has seen remarkable growth over the past decade, especially considering the delivery of goods. Notably, the delivery of medical supplies via Unmanned Aerial Vehicles (UAVs) has emerged as a promising application, which is driven by its societal benefits and has been demonstrated successfully in rural environments. However, expanding these operations to urban areas presents a unique set of challenges, such as high traffic density and the need for online path planning of UAVs. Effective path planning and coordination are of utmost importance to allow for safe and efficient delivery operations in urban environments. Currently, path planning of UAVs is relatively unexplored, and for drone delivery models the trajectories of UAVs are often oversimplified. This research aims to fill the research gap by developing and evaluating an online path planning and coordination mechanism in 3D for a cooperative fleet of autonomous UAVs delivering medical supplies. The Rotterdam Area is selected as a use case, but the model can be adapted to other cities. Utilising an agent-based model formulation, this paper formulates a novel approach for multi-agent pathfinding in 3D and real-world environments, where the travelled paths of UAVs in drone delivery models are computed by taking into account the urban environment and kinematics of UAVs. Windowed Cooperative Safe Interval Path Planning (WC-SIPP) is used as a path planning and coordination mechanism, where kinematic constraints of UAVs and safety separation distances between UAVs are incorporated. Experimental studies show the capability of the algorithm to plan UAVs in different multi-layer urban environments. We establish that including more than one vertical layer is beneficial for UAVs to avoid conflicts by allowing them to change altitudes during flight. Additionally, we show that an increase in the number of layers does not lead to an increase in the delivery time of packages. However, an increase in available layers does come with a decrease in computational performance. Besides, the proposed method is able to plan the trajectories of UAVs and ensure safety separation between UAVs within a dynamic environment, where up to 20 unexpected obstacles are introduced into the environment. Furthermore, we demonstrated that for our operation model, assigning prioritisation based on either mission type or package request time did not influence the delivery time of packages.

With renewed interest in the development of civil supersonic aircraft, their return in the future is becoming more ever more likely. The environmental impact of emissions in the stratosphere on climate and the ozone layer therefore needs to be explored. The stratospheric ozone levels determine the amount of harmful ultraviolet radiation reaching the Earth's surface and thus the level of risk to human health and ecosystems. Ozone response is complex, varying with emission altitude and latitude and we are currently reliant on computationally expensive chemistry-transport models to calculate chemical species concentration changes resulting from supersonic aviation emissions. This paper takes a novel approach to reduce the dependency on these models, creating data-driven dynamical systems that model the global spatiotemporal atmospheric ozone response for different emission scenarios. The dynamic mode decomposition (DMD) and proper orthogonal decomposition (POD) methods are applied to atmospheric ozone data obtained from the GEOS-Chem model, and the evolution of the dominant POD spatial modes are modelled using sparse identification of nonlinear dynamics algorithm (SINDy). We show that DMD models can reconstruct monthly global column ozone changes with root mean square errors less than 0.05 Dobson unit (DU) for a period of three years. Predicting the global mean column ozone changes for the years beyond the period used to construct the models, results in errors less than 0.12 DU. Independent DMD models at two different altitudes can be interpolated to produce estimates for ozone response at an intermediate altitude. These methods can serve as a basis for low dimensional surrogate models that can be used to evaluate chemical species concentrations changes as a result of supersonic aviation emissions.

In this thesis a grey-box model is developed to predict the power increase over time due to marine biofouling. Biofouling is known as the undesired adverse effect of living organisms growing on submerged surfaces. Fouling creates roughness on the hull and propeller and thus additional frictional resistance and loss of propeller efficiency, also referred to as an additional sea margin for ships due to biofouling. The International Maritime Organization (IMO) identified marine biofouling as one of the primary problems from both economic and ecologic points of view. Biofouling threatens the ecological balance of world seas by transferring invasive aquatic species and it causes a reduction in hydrodynamic performance of ships, which in turn increases fuel costs and greenhouse gas (GHG) emissions. A white-box model is first developed to predict biofouling growth together with resulting roughness, and compute increase in frictional resistance, loss of propeller open water efficiency, and change in wave resistance. With this a physical modelled prediction can be made for increase in power. This increase is then used in a data-driven model together with all other used parameters, to improve the prediction and output of the model. With the developed model, relevant questions have been answered from both a research and industry perspective. For Feadship, the developed model was applied for their yachts to give insight in power increase, fuel increase, maintenance increase, speed loss, range loss and added cost due to biofouling. With implementation of the proposed grey box and white box models, predictions can be made for ships varying in all ranges of available data. With an indication of ship profile and parameters, biofouling and its resulting sea margin can be estimated with high accuracy in early stage ship design.

With the increasing availability of sensor data to monitor the health of systems and components, maintenance is encountering a shift from the traditional preventive and corrective methodologies to a predictive approach. Given that maintenance, repair and overhaul are a significant operational cost to organizations, it is imperative that the adoption of predictive maintenance techniques such as prognostics comes to the forefront. Prognostics models ensures that remaining useful life (RUL) estimates are available throughout the life of the component by leveraging sensor data. The obstacle faced by prognostics though, is the lack of standardized metrics, performance evaluation and their application to different case studies. Visualization results in understanding the evolution of the performance over time rather than single point estimates, which helps in identifying performance at crucial points in the life of systems. This paper focuses on assessing and visualizing the performance of prognostics models by addressing the shortcomings of the currently available performance metrics, advancing these metrics to suit the requirements and applying these to visualize the performance of two linear regression models over time. The results show that the modification of currently available performance metrics can enhance performance visualization, leading to better interpretation of the performance of prognostics models.

The adverse human contribution to global climate change has forced the yachting industry to acknowledge the need to reduce its environmental impact due to the client's increasing pressure and potential future regulations to limit the ecological effects. Unfortunately, current real-world data presents a significant disparity between predicted and actual gathered energy consumption results. Thus, this research aims to develop an approach to accurately predict total dynamic Energy Consumption (EC) using real operation voyage data for the improved early-stage design of future yachts. A grey-box modelling (GBM) solution combines: physics-based white-box models (WBM); and black-box model (BBM) artificial neural networks to provide estimations with high accuracy and improved extrapolation capacity. The study utilizes ten months of onboard continuous monitoring data, hindcasted weather, and voyage information from a Feadship fleet yacht. Upon applying a sequential modelling methodology, predictions are compared between the three model categories, which ultimately indicated propulsion and auxiliary estimates fall within 3% and 7% error of operational conditions. The study is then continued using external range datasets to evaluate the extrapolation potential. While GBM improvements are seen over the BBM, limitations were directly related to the strength between dynamic WBM input-output correlations. Finally, a general consideration decision structure is detailed to provide naval architects with practical knowledge and confidence in future applications of the GBM modelling approach and when such methods are appropriate.