http://theses.ncl.ac.uk/jspui/handle/10443/5223 2026-06-24T10:01:33Z 2026-06-24T10:01:33Z Benthem De Grave, Remco Martijn http://theses.ncl.ac.uk/jspui/handle/10443/6815 2026-06-12T14:13:27Z 2025-01-01T00:00:00Z

Title: Designing digital technology to empower climate-sensitive food purchases Authors: Benthem De Grave, Remco Martijn Abstract: A shift towards climate-sustainable diets is essential to achieve climate targets and mitigate disaster risks. Despite many people valuing climate-sensitive behaviours, a lack of knowledge and skill undermines action. Digital technology can aid this transition but faces ethical concerns with persuasive methods that may undermine autonomy. Also, prevalent design decisions may not always align well with actual usage patterns, which may undermine the potential of designs. This thesis examines how to design digital technology that supports climate-sustainable diets while addressing these considerations. The study includes: • A systematic literature review on opportunities and challenges of food purchase choice applications. • Three empirical design chapters detailing participatory input and theory-driven proce dures culminating in ’MyFoodprint’, an app prototype tested in real-world settings. • Atheoretical debate on protecting individual autonomy with behaviour change design. Key insights reveal: • Digital tools should focus on educating users about ’foodprint’ tailored to their purchasing behaviour. • To preserve autonomy, attempts to motivate behaviours directly are best avoided. • Design should not rely on consistent or long-term use. • Positive support for the potential of MyFoodprint to empower sustainable choices. • The use of word clouds visualizing product contributions to one’s overall foodprint ap peared particularly effective alongside support for finding alternative products. • Indirect benefits included sparking discussions around foodprint. iii Contributions include: • Design implications from various study designs, including field observations. • Insights into the practical and ethical feasibility of different design options. • Theory-backed design artifacts available as resources for researchers. • Categorization of intervention techniques based on their alignment with autonomy. This work enhances understanding of how digital technology can empower individuals towards sustainable behaviours without compromising personal autonomy. Description: PhD Thesis

2025-01-01T00:00:00Z Alfagham, Mazyounah Haif S http://theses.ncl.ac.uk/jspui/handle/10443/6812 2026-06-11T15:34:16Z 2025-01-01T00:00:00Z

Title: Attack scenario generator for industrial control system Authors: Alfagham, Mazyounah Haif S Abstract: Attack scenarios are hypothetical or planned sequences of events that describe how an attacker might target a system, organization, or network. Their primary goal is to carry out malicious activity. Attack scenarios help in identifying potential threats and understanding the possible consequences of a successful attack. Security analysts traditionally create attack scenarios manually. They may use graphical security models such as attack graphs, trees, or frameworks such as cyber kill chain or a combination of these. Security analysts heavily rely on their knowledge and experience to carry out this manual approach. However, the manual approach is a challenge for complex systems, such as Industrial Control Systems (ICSs). Indeed, ICSs have various requirements coming from the plurality of structures, devices, protocols and application contexts. In addition, the threat landscape for ICSs is constantly evolving due to their increased use. The manual creation of an attack scenario for a given ICS against a given threat landscape might therefore be complex, error-prone and quickly outdated. The proposed novel general methodology can be effectively used by security analysts to define attack scenarios for ICSs. The proposed methodology gathers the raw data from vast sources to prepare the data and initiate the inferential analysis. Furthermore, it structures and creates the attack sequence to generate the scenario and then simulate the attack scenario. The method was first tested by manually analyzing a complex case study. Human analysts were relied upon to review previous reports and map them with ICS cyber kill chain to generate a scenario and identify the relationship between the attacks. Next, it was demonstrated that this method could be automated. Both a threat-based approach (by automating the cyber threat knowledge base to generate attack scenarios) and a system-based approach (by using the static system state to create attack scenarios) were used. These two approaches were combined in a new tool called the Attack Scenario Generator (ASG). The ASG can generate and optimize attack scenarios based on the cyber kill chain, and predict the techniques, software, and groups behind the attacks in just a few seconds with high accuracy. This saves time, effort, and assists ICS owners efficiently. Description: PhD Thesis

2025-01-01T00:00:00Z Harper, Scott http://theses.ncl.ac.uk/jspui/handle/10443/6802 2026-06-09T10:26:31Z 2025-01-01T00:00:00Z

Title: On the Security and Privacy of Animal Technologies Authors: Harper, Scott Abstract: As the Internet of Things (IoT), smart devices, and their corresponding mobile apps are becoming increasingly widespread, they are expanding into various different industries. One of these rapidly expanding sectors is animal technologies, which includes systems and devices designed to assist with animal care. In pet tech only, it is projected to reach a market value of $3.7 billion by 2026 [104]. However, these systems bring new security, privacy, and safety risks to users, their animals, and their homes. Despite these concerns, the risks of these systems, as well as the users’ apprehensions about these issues, remain under-researched. This lack of research and data protection regulations in this space leaves users vulnerable to attacks and hampers their ability to protect themselves effectively. This PhD work investigates various aspects of the security and privacy of these technologies to inform the current state of risk and user perceptions. Security and Privacy of Animal Apps: In the first part of this thesis, our work involves a range of tools used to perform static, dynamic, network traffic, and privacy policy analysis on a set of 40 animal Android apps (both farm animals and pets). We identify poor security and privacy practices that do not effectively gain the consent of the user and communicate their details in ways that may leave them vulnerable. We additionally find that some of the apps are communicating the user’s login details in plaintext in non-secure http traffic, leaving them vulnerable to very obvious, yet dangerous attacks, by anyone who is able to view this network traffic. These issues were communicated to the companies responsible, with those who responded having the issue fixed upon later retesting. Sensor-based IoT Identification: The second part of the thesis looks at a possible identification method to be used by resource-constrained IoT devices, with limited interaction methods, such as those used on and around animals e.g., at a large scale on a farm. In our proof-of-concept implementation, by utilising the accelerometers already present in such devices (i.e., the Nordic Thingy 52 and 53), we capture the data pattern created from physically tapping two IoT devices together. Our results showcase the feasibility of implementing such a system that is able to correctly identify matching tapping events from IoT devices which want to pair for secure communication. We test a range of signal processing methods, such as the correlation coefficient and energy of the signals, combining those found to be effective for our final similarity calculation. The proposed system is able to achieve an EER of 3.5% when comparing 100 samples of data against each other, with possible adjustments to the threshold to get a lower FAR if needed. User Studies of Animal Technologies Security and Privacy: In this final part of this PhD work, we turn our focus to the users of these systems (more specifically, pet owners). We design a user study in the form of an online survey to understand their views, concerns, and actions regarding these systems. Using Academic Prolific, we distribute this to 593 participants from the UK, US, and Germany (roughly 200 from each) targeting specifically pet owners. This study gives insight into the apps and devices used by pet owners, the perceived advantages and disadvantages, concerns, incidents that have occurred, as well as the different perceptions around the data that may be collected by them and how they might protect themselves from these risks. Despite only a few reported incidents with these technologies, we find 521 of the participants expressed concern about an incident, such as a data leak, and the well-being and safety of their pet. Despite these concerns, these participants took far fewer precautions toward protecting the security and privacy of these systems compared to what they employ for their general online security and privacy. The findings of this PhD research give perspective to the overall security and privacy of animal technologies. Our work contributes to the body of knowledge in a holistic and comprehensive way, i.e., regulations review, system studies, secure system design, and user studies. We provide discussions and recommendations for multiple stakeholders such as academic and industrial researchers and designers, farm owners and managers, policymakers, and the end users of animal technologies. Description: PhD Thesis

2025-01-01T00:00:00Z Kucukgoz, Burak http://theses.ncl.ac.uk/jspui/handle/10443/6774 2026-05-13T11:18:31Z 2025-01-01T00:00:00Z

Title: Enhanced predictive models for macular hole surgery outcomes Authors: Kucukgoz, Burak Abstract: This thesis presents research work conducted in the field of retinal image analysis. More specifi cally, the work is directed at the employment of deep learning (DL) based image informatics for the analysis of diverse real world phenomena where features of interest are very difficult to distinguish. The evaluation of idiopathic full-thickness macular holes (MHs) holds critical clinical importance as MHs represent one of the strongest predictors of surgical success, influ encing both anatomical closure and functional visual recovery– a key motivation for developing robust deep learning frameworks to quantify their characteristics and predict postoperative out comes. In this context, three distinct parts to retinal image analysis are proposed. The first part addresses critical research questions on the quantitative assessment of MH, the role of DL in postoperative visual acuity (VA) prediction, the integration of automated optical coherence tomography (OCT) analysis for clinical decision-making, and the potential of DL models to improve diagnostic accuracy and support clinical practices. Hence, this part presents a compre hensive image informatics framework to create a high-quality spectral-domain OCT (SD-OCT) image dataset, providing a robust DL-based predictive model of VA in patients following surgery with MH and presenting an automated solution for non-standardised SD-OCT datasets. The imaging data undergoes preprocessing, quality assurance, and anomaly detection procedures. Seven state-of-the-art DL predictive models are then designed, implemented, trained, and tested with multiple two-dimensional (2D) input channels on the SD-OCT dataset. The models are quantitatively compared using four evaluation metrics. The method concludes the impact of the following surgery by predicting VA. Overall, the obtained results confirm that the fully automated approach with input from seven central SD-OCT images from each patient may robustly predict VA measurements using a high-quality SD-OCT image dataset. Following this, three-dimensional (3D) convolutional neural networks are integrated to train the model. 3D networks generally outperformed the 2D networks in some evaluation metrics; however, it came with the sacrifice of significantly more computational complexity. The second part identifies key research questions related to common sources of uncertainty in OCT images and proposes an effective method for representing and quantifying this uncertainty in DL-based predictive models. Furthermore, the study compares the proposed UQ method with existing approaches. In this context, the study highlights the significance of uncertainty, especially in dealing with the SD-OCT images. Predicting postoperative VA through DL models is crucial for decision-making and patient advisement, though their black-box behaviour is opaque to users and uncertainty associated with their predictions is not typically stated, leading to a lack of trust among clinicians and patients. To meet this need, an uncertainty-aware regression model is introduced for predicting postoperative VA using 3D SD-OCT images. The model not only x predicts VA post-surgery but also quantifies the associated uncertainty, enhancing reliability and trustworthiness. Qualitative evaluation shows that the proposed model outperforms commonly used methods in terms of prediction accuracy and reliability, demonstrating robust performance on out-of-sample data, including low-quality images and previously unseen instances. This makes the model a promising tool for clinical settings, improving the reliability of DL models in predicting VA. The third is the segmentation of the retinal external limiting membrane layer, where any disruptions in this layer are associated with worse visual outcomes in patients with idiopathic full-thickness MHs. Precise image-wise binary annotations are used to segment the retinal external limiting membrane (ELM) layer. Finally, qualitative and quantitative results are systematically compared with seven state-of-the-art DL-based segmentation methods to identify the ELM layer with an automated system. Additionally, it examines the feasibility of integrating automated ELM layer segmentation into clinical workflows while incorporating the latest advancements in DL-based ELM detection. The results confirm the efficacy of DL in retinal image analysis, providing a foundation for future enhancements in clinical applications. Future work will explore enhancing the models’ performance and efficiency, and extending the approach to other retinal conditions. Keywords: Image Analysis, Machine Learning, Deep Learning, Visual Acuity Measurement, Optical Coherence Tomography Description: Ph. D. Thesis.

2025-01-01T00:00:00Z