https://www.um.edu.mt/library/oar/handle/123456789/129100 2026-04-05T12:28:47Z https://www.um.edu.mt/library/oar/handle/123456789/144767 Title: A ‘prosthesis life-cycle ontology’-based service system framework to cater for amputees’ evolving needs Abstract: Lower-limb prostheses remain essential to restoring mobility and independence for amputees, yet their design and aftercare continue to be fragmented, reactive, and inefficient. Current practices are characterised by siloed stakeholder input, tacit rather than structured knowledge, and limited methods for adapting to the evolving needs of both amputees and prosthesis sub-systems. This thesis addresses these systemic inefficiencies by developing the Adaptive Prosthesis Life-Cycle Service System Framework (adProLiSS), an ontology-based, consequence-aware framework that reconceptualises prosthesis management as an integrated product–service system (PSS). The research aim was to improve amputees’ overall prosthesis experience, measured in terms of time efficiency, cost efficiency, functionality, user comfort, emotional well-being, and aftercare quality, through the design, implementation, and evaluation of a prosthesis life-cycle framework that supports adaptive, knowledge-based decision-making. To test this aim, seven research questions (RQ1–RQ7) were formulated, spanning the identification of life-cycle activities, stakeholder roles, tools and methods, critical constraints, specifications of a prosthesis PSS, and its applicability and evaluability across prosthesis types. Conceptually, the thesis advances the state of the art by introducing the Prosthesis Life-Cycle Consequence Knowledge Modelling Frame (PLCCKMF), a domain-specific ontology that formalises heterogeneous consequence knowledge, including intended and unintended, interacting and non-interacting outcomes, across physical, functional, emotional, systemic, and semantic domains. Methodologically, the work contributes a closed-loop life-cycle methodology, digital tool support through the Patient–Prosthesis Management System (PPMS) and the Consequence-Driven Co-Design Support Tool (CD-CST), and a scenario-based evaluation strategy combining physical demonstrators, digital platforms, and stakeholder engagement. Practically, these contributions were validated through prototype demonstrations addressing ulcer detection, weight distribution monitoring, fall risk, maintenance alerts, and daily logging. Evaluation findings indicate that adProLiSS delivers measurable improvements across the six evaluation criteria, enhances stakeholder collaboration, and enables decision traceability in ways not achieved by conventional approaches. The framework was shown to reframe amputees from passive recipients to active contributors of consequence knowledge, support clinicians in preventive and evidence-based aftercare, and allow engineers to anticipate design consequences before physical prototyping. The findings carry implications for service providers and policymakers, particularly in highlighting how consequence awareness and interoperability standards such as HL7/FHIR could guide future procurement practices and rehabilitation service models. The research is not without limitations. The scope was bounded to lower-limb prostheses, early-stage digital and physical prototypes, and a geographically localised but scientifically justified stakeholder sample. Nonetheless, the framework establishes a demonstrable foundation that can be extended to upper-limb prostheses, scaled through multi-centre evaluations, and advanced via ontology governance methods and higher-TRL digital systems. Future work also envisions adProLiSS as an educational simulator and as a bi-directional communication platform for real-time patient–clinician interaction. In conclusion, this thesis contributes to engineering design knowledge by embedding consequence awareness and ontology-driven reasoning into the prosthesis life-cycle, demonstrating how structured knowledge representation and adaptive service-system integration can improve both design and aftercare. The adProLiSS framework thus stands as a novel paradigm for prosthesis life-cycle management, bridging conceptual, methodological, and practical advances, and as a commitment to enhancing the lived experiences of amputees through more collaborative, knowledge-based, and adaptive healthcare systems. Description: Ph.D.(Melit.) 2025-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/144766 Title: Graphene oxide on porous polyurethane substrates : synthesis, deposition, and characterisation Abstract: Graphene oxide (GO) is a chemically functionalised, two-dimensional nanomaterial with broad applicability in areas such as water purification, sensing, coating and membrane fabrication. Its abundant oxygen-containing functional groups enable strong interactions with various polymers, making it a suitable candidate for forming multifunctional hybrid materials. Traditional GO synthesis methods involve corrosive acids, strong oxidising agents, and elevated temperatures, which pose safety risks due to the release of toxic fumes and the potential of forming explosive by-products. This doctoral study proposes a simplified, room temperature method using only concentrated H2SO4 and KMnO4, eliminating the need for heating and reducing hazardous emissions. Characterisation confirmed the effective oxidation and exfoliation of graphite with an average lateral flake size of 15.06 ± 1.87 μm and 87 layers, into GO sheets with resultant lateral sheet size of 10.22 ± 0.62 μm and ~ 10 layers. X-ray Photoelectron Spectroscopy (XPS) analysis showed that in-house synthesised GO had a comparable oxygen content and functional group distribution to commercial GO, with epoxy (C–O) as the dominant group. X-ray induced Auger spectroscopy revealed a reduction in sp2-hybridised carbon from 80.21% in graphite to 45.52% in GO, confirming effective oxidation. The second phase of this study focused on the deposition of the synthesised GO on the top surface of a modified porous polyurethane (PU) support. Due to the inherent chemical inertness of PU, surface modification was carried out using chromic acid etching followed by polydopamine (PDA) deposition at a fixed concentration and varied deposition times (6, 12, and 24 hours). Characterisation showed that chromic acid altered the PU surface both physically and chemically, increasing hydrophilicity. PDA further modified the surface, introducing functional groups such as COOH/N-C=O, C-O-C=O, and C=O, thus enhanced the compatibility with GO. GO was then deposited at two concentrations (1 g/L and 10 g/L) and three different deposition times (6, 12, and 24 hours). Only the higher concentration resulted in uniform coverage. Characterisation confirmed the successful chemical anchoring of GO to the PDA-modified PU, facilitated by reactions between the oxygen functionalities in GO (acid anhydride and epoxy groups) and the amine groups in the PDA layer. Description: Ph.D.(Melit.) 2025-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/144354 Title: Data analytics for photovoltaics reliability and operations Abstract: The rapid growth of photovoltaic (PV) systems has increased the importance of reliable and standardised monitoring frameworks to ensure sustained performance and operational reliability, particularly in challenging environments such as small island states. This research was conducted within the framework of the Horizon Europe PROMISE project, which brought together an interdisciplinary consortium of scientists and engineers to investigate innovative approaches to PV system reliability, digitalisation, and sustainable operation through the use of Living Laboratories and Test Sites. The primary aim of this MSc by Research study was to design, implement, and validate a PV monitoring and anomaly detection framework aligned with internationally recognised standards and open-data practices. A comprehensive review of IEC 61724 and related literature was first undertaken to establish best practices in PV performance monitoring, data quality assurance, and anomaly detection. This review informed the system architecture and provided a reference framework for both the present work and future research activities. The experimental implementation involved the deployment of electrical and meteorological monitoring hardware across ten PV Living Laboratories in Malta. Sensors were selected and justified based on standard compliance and site-specific constraints and were integrated using local controllers and standardised communication protocols. Data acquisition was performed locally using Raspberry Pi devices, while a scalable cloud-based architecture was developed using Microsoft Azure to support data ingestion, storage, processing, and real-time visualisation through Power BI. Anomaly detection was implemented in two stages, addressing both low-level operational faults and higher-level performance deviations. The framework was validated using real operational data from the Malta PV Living Laboratories, demonstrating its ability to improve system observability and support early fault identification. The outcomes of this study highlight the suitability of Living Laboratories for applied PV research and position Malta as a practical testbed for the development and evaluation of digitalised PV monitoring solutions relevant to wider European contexts. Description: M.Sc.(Melit.) 2025-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/143783 Title: Examining the effects of different winglet designs on aerodynamic forces using computational fluid dynamics Abstract: The pursuit of better aerodynamic efficiency in aircraft design has led to significant advancements, with winglets being one of the most impactful innovations. Winglets are aerodynamic extensions at the tips of an aircraft’s wings designed to reduce induced drag by minimizing wingtip vortices. This reduction in drag improves lift-to-drag ratios, leading to fuel savings, extended range, lower carbon emissions, and quieter flight operations. Over time, various winglet designs have emerged, including bio-inspired configurations, each offering distinct aerodynamic advantages. While traditional wind tunnel tests remain useful, Computational Fluid Dynamics (CFD) has become essential for optimizing winglet geometry, providing insights into flow fields and aerodynamic forces under various conditions. Despite progress, designing universally optimal winglets remains challenging due to the need to balance lift, drag, structural strength, and manufacturing feasibility. This thesis uses CFD to examine the effects of different winglet geometries on aerodynamic performance, contributing to a deeper understanding of their impact on efficiency. First, an intensive review of the current literature related to CFD studies of winglets was carried out. The different setups used in literature combined with further information obtained through background theory research were used to determine the set up used for the simulations carried out in this thesis. Within this study, no winglet, raked winglet, fenced winglet, blended winglets and split winglet models were analysed. In this thesis, the end section of the winglet attached to the full wing geometry and the end geometry by themselves were analysed and compared. These results provided similar trends for both geometries, with the raked winglet providing the highest lift-to-drag ratio, followed closely by the blended winglet. From this study, the split and fenced winglet provided the lowest values of lift-to-drag ratio meaning that these winglets were the least efficient. Furthermore, to analyse clearly the trailing vortices formed behind the winglet, another set of simulations were carried out with the same geometry and domain, only editing the mesh to be more refined behind the wing. This resulted in a more clear display of the vortices, making it easier to compare how the different winglets effected the formation of these vortices. The effect of the Cant angle on blended winglets was also analysed, to see which model produced the best lift-to-drag ratio. The results were very close to each other. The 45° Cant angle winglet produced the best results, closely followed by the 30° Cant angle model and lastly by the 60° Cant angle model. Description: B.Eng. (Hons)(Melit.) 2025-01-01T00:00:00Z