Electronic Theses and Dissertations (ETD)

Permanent URI for this collection

https://dspace.library.uvic.ca/handle/1828/12

For information on how to submit your thesis to this collection, please go to our UVicSpace Guidelines.

Access to the full text of some theses may be restricted at the request of the author.

All theses from 2011 to the present are in this collection, as well as some from 2010 and earlier years.

Browse

Now showing 1 - 20 of 11856

Mechanochemical fabrication of ZnO nanoparticle inks for eco-friendly low-cost thin film environmental sensors
(2026) Boakye, Gibson Asumani; Papadopoulos, Christo
The increasing demand for low-cost, scalable, and energy-efficient sensing technologies has driven significant interest in nanomaterial-based thin film sensors. This thesis presents the fabrication and characterization of zinc oxide (ZnO) nanoinks via a low processing temperature mechanochemical planetary ball milling (PBM) approach for the fabrication of eco-friendly, flexible, and low-cost gas sensors. The study focuses on establishing a low input energy, simple and scalable solution-based process that enables the direct conversion of bulk ZnO powder into functional nanoinks suitable for thin film deposition. ZnO nanoinks were prepared through wet ball milling under varying conditions, including milling speed 200 rpm to 1000 rpm, time 10 minutes to 120 minutes, and solvent type (DI water, ethylene glycol and Isopropyl alcohol), to investigate their influence on nanoparticle size, morphology, and dispersion. Thin films were subsequently fabricated by deposited the nanoinks onto a wide range of low-cost substrates including glass slide, filter, plain and lined paper, plastic polymer, foil, ceramic and flexible materials using an adjustable blade applicator technique to form films with controlled thickness (15 μm to 50 μm) and uniformity. The versatility of substrate selection highlights the potential for low-cost and flexible sensor fabrication. Material characterization was conducted using techniques such as scanning electron microscopy (SEM), transmission electron microscopy (TEM), atomic force microscopy (AFM), Raman spectroscopy, ultraviolet–visible (UV–Vis) spectroscopy, and energy dispersive X-ray spectroscopy (EDX), confirming the formation of nanoscale ZnO. Variations in milling conditions were observed to affect particle size, dispersion, and film morphology. Initial gas sensing measurements of the fabricated ZnO thin films on different substrates was evaluated under various conditions, including different gas species (Hydrogen, dry air and argon), gas concentrations (low and high), and different flow systems (static and continuous flow) at room temperature. The sensors demonstrated measurable and reproducible responses across both low and high gas concentrations, with response magnitude increasing as gas concentration increased. These results represent a proof-of-concept for room-temperature sensing using solution-processed ZnO films. The data indicated that sensors milled at 200 rpm had response times of 610–750 seconds under static flow, and 750 rpm sample reaches about 2100 seconds under continuous flow. Recovery times were generally faster with less variations, 560–660 seconds for most samples. Glass-based films generally exhibited quicker response and recovery kinetics than paper-based substrates. The sensing mechanism may be attributed to surface adsorption and desorption processes involving oxygen species, which modulate charge carrier concentration and electrical conductivity upon exposure to target gases. This work demonstrates that PBM-fabricated ZnO nanoinks enable low-cost, scalable fabrication of flexible thin film gas sensors. The preliminary gas sensing results provide insight into how processing conditions influence material properties and sensing behavior, highlighting potential for environmental, industrial, and wearable sensing applications.
Genetic and genomic characteristics of a selfish X chromosome in Drosophila testacea
(2026) Clark, Michael W.; Perlman, Steven John
Meiotic drivers are selfish genetic elements that manipulate the process of meiosis to increase their transmission, circumventing Mendel’s Law of equal segregation. Because meiosis fundamentally differs between the sexes, meiotic drivers use different mechanisms and strategies in male and female meiosis. While drivers in male meiosis use a strategy of killing meiotic products which don’t contain them, drivers in female meiosis will utilize a strategy of altering their segregation to ensuring a “winning” orientation. The mechanism that female meiotic drivers use to alter segregation is not well understood. For this thesis, I investigated a selfish X chromosome that causes both male and female meiotic drive in the woodland fly Drosophila testacea. In the first chapter, I provide an overview of meiotic drive, providing more background on the focus of my thesis, female meiotic drive. In the second chapter, I used controlled crosses to test whether one or two copies of this selfish X chromosome cause higher rates of nondisjunction, or chromosome segregation failure, in female flies. I found that females carrying one selfish X chromosome failed to transmit either of their X chromosomes at a significantly greater rate than wildtype females; this was even more pronounced in females with two selfish X chromosomes, creating a large number of sterile sons with an XO genotype. In contrast, females carrying two selfish X chromosomes almost never transmitted both of their X chromosomes. These results suggest that the two selfish X chromosomes interfere with each other over segregation in homozygous D. testacea females. Meiotic drivers tend to be structurally distinct from their non-driving counterparts, often containing chromosomal rearrangements like inversions or expansions of repetitive elements. In the third chapter, I analyze the selfish X chromosome for such structural differences, using Hi-C sequencing reads to generate chromosome-level scaffolds of the wildtype and driving X chromosomes. In doing so I demonstrate that the driving X chromosome has acquired an almost complete Wolbachia genome. Wolbachia are a group of intracellular bacteria symbionts found in insects and other terrestrial arthropods, notable for their ability to manipulate host reproduction to increase their transmission. Through comparisons with several Wolbachia genomes, I characterize the Wolbachia insertion and speculate on possible functions it may have. The work performed in my thesis helps to characterize some of the fascinating genetic and genomic characteristics of the selfish X chromosome, opening the door to further questions and research about female meiotic drive.
Systemic drivers of state formation
(2026) Iqbal, Husnain; Watson, Scott D.; Ramraj, Victor V.
This study develops a systemic theory of state formation that places dynamic change in international systems, rather than the bellicist theory’s emphasis on war occurrence, at the centre of long-term political development. War-centric scholarship has illuminated important aspects of state formation in Europe, but its explanatory reach narrows when extended beyond that historical context. The literature on non-European state formation is divided between internalist and externalist accounts that privilege a single scale of analysis and often rest on short time horizons that obscure longer-run sequences of consolidation and reversal, producing mixed findings and a fragmented scholarly consensus. Moreover, where external dynamics are considered, international politics is often treated as fixed and divided along a West/non-West binary. These limitations produce problematic implications: non-Western societies are portrayed as inherently incompatible with modern statehood or as destined for enduring weakness, while relatively strong cases of extra-European state formation remain unexplained. These problems stem from the incorporation of the contentious ‘decline-of-war’ thesis, a partial account of hierarchy, and, most critically, a neglect of systemic change in international politics. I address these limitations by integrating insights from International Relations theory, historical sociology, global history, and political geography into a framework that explains how states emerge, transform, and persist across time and space. My core claim is that state formation is fundamentally an international-systemic process and is best understood in these terms rather than in monadic or dyadic ones. Shifts in system-wide capabilities periodically reconfigure the international-political setting within which political authority is organised, territorially anchored, and spatially extended. Political units respond to these shifts through material, institutional, and ideational practices of spatial ordering. The timing and intensity of state formation are conditioned by systemic-contextual influences conceptualised as ‘structural modifiers’, including geography, technological change, transnational networks, and shifts in the forces of production and destruction that shape strategic exposure and organisational capacity. Great powers are principal actors in this process because their ordering strategies translate systemic pressures into uneven effects on spatial transformation across multiple scales, extending into national and subnational arenas, including ‘grey spaces’ where formal sovereignty and effective control diverge. These strategies are mediated by the responses of other political units and by social forces. State formation is therefore a multiscalar process that unfolds through interactions among political units as systemic conditions shift over time. Empirically, I develop a longue durée account of state formation across Eurasia to explain why the national state consolidated earliest in Europe despite parallel islands of territorial consolidation elsewhere. I also show how the national state form drew on polycentric foundations, with key practices and institutional templates developing earlier or in parallel across multiple Eurasian systems before consolidating unevenly into a dominant model across the emergent global system. The empirical chapters then apply this framework to South Asia and Southeast Asia, organised around major river systems as spatial frames. Across these cases, the findings support the claim that state formation remains historically open and spatially uneven. Even without system-wide war since 1945, sustained competitive pressures continue to drive spatial transformation through alignments, economic statecraft, infrastructural integration, and the extension of security architectures against emergent threats.
Leveraging local wisdom: The role of community knowledge in the strategic direction of BC’s community foundations
(2026) Tom, Vincent Eric; Thiessen, Susanne
Community foundations (CFs) in British Columbia are uniquely positioned as place-based philanthropic institutions whose mandates depend on their ability to understand and act on community knowledge (CK). Yet how senior CF leaders conceptualize CK, access it, and mobilize it through organizational decision-making structures remains under-researched in the Canadian philanthropic literature. Using a constructivist grounded theory methodology, this study conducted ten semi-structured interviews with senior leaders across established BCCFs to examine how CK informs strategic direction and decision-making. The analysis of these interviews produced five categories describing CK’s movement through the CK: cultivating informal access, managing multiple knowledge systems, navigating board dynamics, asserting strategic agency, and returning knowledge through reciprocity. Bringing together these categories supports an emerging theory, the Stewardship of Relational Knowledge, which articulates CK mobilization as a cyclical, non-linear process of stewardship conditioned by institutional power. The study concludes that a CF’s legitimacy as a community actor is not granted by its endowment size or grant programs but earned through its stewardship of the knowledge entrusted to it by the community.
Understanding the percieved relationship between physical activity and mental health in adolescents: A follow-up to the Adolescents' Daily Lives (ADL) study
(2026) Grant, Charlotte Grace Noelle; Buckler, Jean
Adolescence is a critical stage of human development when lifelong patterns of physical and mental health begin to take shape. Current adolescents’ levels of physical activity and mental health are lower than those of prior generations. Existing research, primarily quantitative in nature, has identified strong associations between physical activity and improved mental health. Though few studies have employed a qualitative design to incorporate adolescents’ perspectives into the robust body of evidence. The primary objective of this study was to explore the barriers and facilitators that influence physical activity engagement among adolescents who report symptoms of anxiety and depression and who do not meet the physical activity recommendations based on the Canadian 24-Hour Movement Guidelines. The sample was recruited through a secondary analysis of data from the Adolescents’ Daily Lives (ADL) study. The ADL study observed adolescents’ daily movement and mental health through wearable devices and baseline and intensive longitudinal survey data. Using a qualitative descriptive design for the current study, semi-structured interviews were conducted with a subset of participants (N= 13) experiencing both elevated mental health symptoms and low physical activity levels. Through reflexive thematic analysis, three themes were generated: (1) intrapersonal barriers, exploring participants’ mental health and daily routines, (2) interpersonal barriers, including coach and team dynamics, and (3) facilitators to physical activity engagement, such as social support and a sense of enjoyment. Participants described their perception of physical activity and mental health as bidirectional, with each one impacting the other in their daily lives. Findings highlight the importance of implementing youth-centred, autonomy-supportive, and socially inclusive physical activity environments to improve adolescent engagement in physical activity.
Investigating tumor metabolism with mass spectrometry imaging
(2026) Shiyuk, Madison; Goodlett, David R.
Tumor metabolism plays a critical role in shaping the tumor immune microenvironment and contributes to immunosuppression in cancer. In ovarian cancer, metabolic competition and accumulation of immunoregulatory metabolites can impair T cell activation and promote exhaustion, resulting in immunotherapeutic resistance and poor patient outcomes. These tumors are highly heterogeneous, and metabolic interactions between cancer cells and immune cells often occur within spatially restricted microenvironments. Thus, mass spectrometry imaging (MSI) has emerged as a powerful technique for spatial metabolomics, enabling label-free mapping of metabolites and lipids directly within tissue sections while preserving tissue organization. However, several analytical limitations remain for spatial profiling of tumor immune microenvironments. Small molecules, such as methionine cycle metabolites, are implicated in immune modulation in tumors, yet they remain difficult to detect using conventional approaches for matrix-assisted laser desorption/ ionization mass spectrometry imaging (MALDI MSI). Further, use of a single MSI modality results in inherent specificity for target molecular classes, motivating development of multimodal workflows integrating two or more MSI approaches. The objective of this work was to develop and apply MSI-based workflows to investigate metabolic heterogeneity within the tumor immune environment. Methodological approaches were established for spatial detection of methionine cycle metabolites using MALDI MSI with on-tissue chemical derivatization. A complementary multimodal workflow was developed to support multimodal imaging experiments combining metabolite, lipid, and immune marker detection within tumor tissues. Application of multimodal workflows revealed spatial metabolic differences associated with immune cell infiltration in murine tumor models and human ovarian cancer specimens. Regions enriched for CD8+ T cells exhibited distinct metabolic signatures, including increased abundance of several amino acids and lipids containing arachidonic acid fatty acyl chains. Together, this work demonstrates the utility of mass spectrometry imaging for spatially resolving metabolic features of the tumor immune microenvironment and provides technical foundations for integrating metabolite, lipid, and protein imaging to study immune-metabolic interactions in cancer.
Evaluation of gold nanoparticles as radiosensitizing agents for high dose rate brachytherapy
(2026) Cecchi, Daniel; Chithrani, Devika
Radiation therapy (RT) is a pivotal part of more than half of all cancer patients’ treatment plan. Despite its clinical effectiveness, normal tissue toxicity remains a fundamental limitation, restricting dose escalation to the tumour and constraining tumour control. Radiosensitizers like gold nanoparticles (GNPs) offer a potential strategy to locally enhance dose deposition in tumour tissue. Due to their high atomic number, GNP-mediated radiosensitization is pronounced at lower photon energies, where increased photoelectric interactions lead to a greater generation of secondary, cell-damaging species. Compared to high-energy external beam radiotherapy (EBRT) (≈1-10 MeV), high dose rate brachytherapy (HDR-BT), which employs radioactive sources emitting low to intermediate-energy photons (≈ 0.1-0.5 MeV), may be an ideal candidate for the introduction of GNPs as radiosensitizing agents. However, preclinical investigations of GNP-induced radiosensitization in HDR-BT remains scarce, largely due to technical challenges of dose delivery in vitro and in vivo due to steep dose gradients and complex treatment delivery. This dissertation addresses these limitations by developing novel irradiation platforms to facilitate robust pre-clinical investigations of GNP-mediated radiosensitization in HDR-BT. We first developed a novel Solid Water plastic phantom to enable uniform HDR-BT irradiation to the base of a petri dish containing cell cultures. Dosimetric verification demonstrated accurate dose delivery and acceptable uniformity across the petri dish within ±5%. Using this platform, GNP-mediated radiosensitization was evaluated during irradiations from both HDR-BT and EBRT to assess energy dependent differences in radiosensitizing efficacy. We found that under HDR-BT, GNPs elicit effective radiosensitization, evidenced by an increase in 53BP1-related DNA damage foci between 30-100% and a marked reduction in clonogenic survival between 2-5%. When the delivered dose rate was matched between HDR-BT and EBRT irradiations, comparable levels of GNP-induced radiosensitization were observed, indicating a dose-rate-dependent component independent of beam energy. In addition, radioactive decay during the clinical lifetime of HDR-BT sources was found to significantly reduce GNP radiosensitization, with the clonogenic survival decrease diminishing from ≈15% with a high-strength source to no measurable enhancement during low source-strength radiations. Next, we optimized the route of administration and functionalization strategy to maximize tumour accumulation and retention in vivo. Intratumoural (i.t.) uptake iv of PEGylated and RGD-functionalized GNPs were investigated following intravenous (i.v.) and i.t. delivery. Both PEGylation alone and RGD functionalization resulted in poor tumour accumulation after i.v. injection but high tumour accumulation and retention following i.t. administration, indicating the viability of either functionalization strategy to be retained within the tumour tissue post-i.t. injections. This work supports the feasibility of i.t. GNP delivery in future HDR-BT workflows. Finally, in vivo irradiations were accomplished using a novel 3-D printed, non invasive irradiation jig designed to encompass the lateral side of tumour-bearing mice. Dosimetric verification revealed accurate tumour dose delivery with minimal toxicity concerns to tumour-bearing mice. PEGylated and RGD functionalized GNPs were subsequently investigated following i.t. administration and compared to in vitro radiosensitization outcomes. Consistent across both in vitro and in vivo experiments, PEGylated GNPs did not produce measurable radiosensitization. In contrast, RGD functionalized GNPs, which promotes intracellular uptake with active receptor targeting, resulted in a significant 20% increase in DNA damage, 20% reduction in clonogenic survival, 57% stunted growth in tumour spheroids, and a pronounced 50% tumour growth delay following HDR-BT relative to control. These findings demonstrate the efficacy of targeted GNP radiosensitization in HDR-BT and demands further investigation into their translational safety and toxicity profiles.
Beyond the classroom: A critical understanding of community-based sex education
(2026) Merritt, Krys; McGregor, Catherine; Clover, Darlene E.
Community-based organizations, which operate within the context of marginalized populations, are well-positioned to provide comprehensive sex education to individuals who are often underserved. Comprehensive sex education integrates the social dimensions of sexuality by addressing individual, interpersonal, community, and societal factors that shape sexual health. In contrast, school-based sex education is inconsistently implemented, resulting in inadequate sex literacy, particularly among those for whom such education is inaccessible or unrealistic. Marginalized communities are systemically placed at greater risk for negative sexual health outcomes and are less likely to access relevant sex education. This study sought to understand the justifications, perceived effectiveness, and discourses surrounding sex education programs offered as community services. To achieve this, sex educators working in community-based organizations were interviewed, and critical discourse analysis was applied to examine how community-based sex education can either disrupt or reinforce the power dynamics underlying inequities among vulnerable populations. The findings indicate that these educators identified dominant discourses influencing their work and described how their practices aim to address these issues by prioritizing collective action, honouring lived experiences, fostering agency, and explicitly challenging normative definitions of sex and sex education.
Transcutaneous vagus nerve stimulation during pregnancy alters the behavioural developmental trajectory and adult microglial phenotype in healthy and prenatally immune-challenged male offspring
(2026) Gargus, Makenna; Tremblay, Marie-Ève
Maternal immune activation (MIA) is the maternal exposure to immunogens and stressors that leads to an elevated maternal pro-inflammatory immune response, which has been linked to a significantly increased risk for offspring neurodevelopmental disorders, such as schizophrenia and autism spectrum disorder. The key links in the mechanism of MIA are the maternal cytokines produced in response to stressors, specifically interleukin (IL)-6. This cytokine penetrates the placenta and the fetal brain, where it binds to microglia, the innate immune cells of the central nervous system, to upregulate pro-inflammatory immune responses. In this thesis, I investigated the potential of vagus nerve stimulation (VNS), a non-pharmaceutical treatment strategy shown to downregulate peripheral immune responses, as a maternal treatment to prevent excessive pro-inflammatory immune responses and alter MIA-related changes to offspring neurodevelopment. Pregnant C57BL/6J mice were treated with VNS from embryonic day (E)15 to E19, while MIA was modelled by intraperitoneal injection of the viral mimetic polyinosinic-polycytidylic acid (polyI:C) on E17. The consequences on maternal cytokines were characterized 3 hours after injection. Offspring were tested for behavioural alterations across development at stages corresponding to juvenile, adolescence, and adulthood, using the open field test, novel object recognition test, and pre-pulse inhibition test, to measure alterations to anxiety and locomotion, recognition memory, and sensorimotor gating, respectively. The hippocampus, a key region of dysregulation in both neurodevelopmental disorders and in MIA models, was chosen in adult offspring for cellular analysis focused on microglia. Varying facets of the behavioural developmental trajectory were altered in male offspring by VNS and MIA. These behavioural changes were accompanied by pronounced changes in microglial morphology in the hippocampus, indicating MIA significantly reduced microglial size and complexity, which maternal VNS exposure could not prevent. Additionally, VNS in healthy dams seemed to increase microglial phagocytic activity in offspring compared to the offspring untreated healthy dams, while VNS in MIA-challenged prenatal development decreased phagocytosis. By contrast, in females we saw minimal behavioural alterations and no significant alterations to offspring adult microglial populations, indicating an important sexual dimorphism linked to the effects of both MIA and VNS. Overall, I characterized maternal immune responses and the resulting influence this had on offspring behavioural and neurodevelopmental outcomes, which were shaped by maternal VNS treatment in health and immune challenge, indicating the importance of further analyzing the mechanisms of both and their unique interaction in offspring neurodevelopment.
Examining the role of appraisal of task value and task understanding practices on the perceived effectiveness of emotion regulation in a high-stakes, collaborative academic context
(2026) Munoz Bocanegra, Yiseth Bibiana; Miller, Mariel
Emotion regulation constitutes a fundamental dimension of self-regulated learning (SRL), particularly within collaborative environments where socio-emotional challenges frequently arise. While preparatory practices such as Task Understanding and Appraisal of Task Value are believed to influence learners’ evaluations of their emotion regulation, the extent to which these practices jointly predict perceived effectiveness in collaborative achievement remains understudied. This investigation examined the extent to which these two crucial SRL practices, task Understanding practices (TUP) and appraisal of task value practices (ATVP), predict students' perceived e emotion regulation (ER) effectiveness during a negative experience in a high-stakes, collaborative case-based performance assessment (CCPA). Using a cross-sectional design, 137 undergraduates reported their TUP and ATVP via self-report measures and evaluated the effectiveness of their ER strategies using a narrative response construction tool. Results revealed that task understanding practices did not show a simple linear relationship with perceived effectiveness, suggesting a more complex or nonlinear relationship that warrants further investigation of potential non-monotonic effects. ATVP demonstrated a positive association with perceived ER effectiveness (OR = 2.10, p = 0.058 in the reduced sample after removing influential outliers). Although this was not statistically significant at the .05 threshold, the effect size is theoretically relevant, indicating that each unit increase in ATVP nearly doubles the likelihood of perceiving the ER as effective. Overall, the combined influence of TUP and ATVP accounted for only a limited proportion of the variance in the outcome, consistent with the theoretical framing of TUP and ATVP as preparatory conditions that shape evaluative standards, rather than as direct causes of regulatory success. These findings challenge linear models of SRL, underscoring the complex roles that cognitive clarity and motivational value play in the perception of ER effectiveness. TUP may be advantageous only when it fosters coherent understanding, whereas ATVP may serve as a motivational anchor during socio-emotional challenges. These insights offer a nuanced perspective on the cognitive and motivational foundations underpinning regulatory evaluation in collaborative learning.
Toward interoperability between FEM systems: A domain-aware framework for translating and benchmarking FEBio models in PolyFEM
(2026) Pak, Zahra; Schneider, Teseo
Finite element interoperability remains difficult when simulation models must be transferred between systems with different input structures, material semantics, and physics support. This thesis inves- tigates that problem through a case study of translating FEBio models into PolyFEM-compatible representations. The thesis presents a modular Python framework for parsing, translating, and bench- marking FEBio models in PolyFEM. The framework includes a FEBio parser, an intermediate repre- sentation, a translation layer for generating PolyFEM-compatible artifacts, and an automated bench- mark runner with PolyFEM execution validation. Each benchmark case is classified as success, par- tial, unsupported, or failed, allowing interoperability limitations to be analyzed systematically. The framework was evaluated on 126 FEBio benchmark cases. Of these, 26 converted successfully, 20 converted partially using documented material fallbacks, 78 were classified as unsupported because the required physics formulations are not implemented in PolyFEM, and 2 failed due to parser- level errors. Among the 46 cases where a sim.json file was generated, 39 executed successfully in PolyFEM, giving an execution success rate of 84.8% . The results show that the primary interoperability barrier between FEBio and PolyFEM is physics scope rather than XML-to-JSON syntax translation. In particular, fluid-structure interaction, biphasic poroelasticity, and reaction-diffusion models could not be translated because PolyFEM lacks corre- sponding formulations. Within PolyFEM’s supported solid-mechanics scope, the framework achieved reliable structural conversion and high execution success. This thesis contributes a working FEBio- to-PolyFEM interoperability framework, a four-status benchmark methodology, and an empirical analysis of structural and semantic transferability between the two FEM systems.
Object-wise metric distance estimation from a single RGB image via semantic and geometric reasoning
(2026) Sultana, Abida; Yang, Hong-Chuan
Estimating metric object distance from a single RGB image is challenging because monocular depth does not provide an absolute scale. Existing solutions either require active sensors such as LiDAR or stereo, rely on monocular depth that remains scale-ambiguous, or use implicit vision-language reasoning that can be unstable for precise measurement. This thesis proposes a semantic–geometric pipeline for recovering metric scale by combining open-vocabulary object grounding and segmentation, label normalization, monocular depth, and camera cues. Object-centric 3D points are reconstructed from the predicted depth, an oriented 3D bounding box is fitted to estimate object dimensions, and real-world size priors are used to compute a scale factor that converts relative depth into absolute distance. The proposed method is evaluated on HOT3D, ScanNet, ARKitScenes, and a custom iPhone dataset, achieving Multi-Threshold Relative Accuracy (MRA) values of 68.85%, 88.30%, 75.12%, and 89.85%, respectively, under the per-frame average mean-distance strategy. The results show that frame-level averaging improves stability by reducing the influence of instance-level outliers. The main limitations of the approach are its dependence on segmentation and depth quality, sensitivity to canonical size priors for categories with high size variation, possible instability under occlusion or truncation, and relatively high processing time. Future work includes more robust scale estimation, adaptive size priors, improved object fitting, the use of consecutive frames for temporal consistency, and pipeline optimization for lower latency.
Mnidoo-mkwendamwin: Beading and restitching with ancestral threads of memory
(2026) Whetung, Estrella; Walsh, Andrea N.
This work was created to go beyond the study cultural mnemonic devices and into the realm of documenting how to make ancestral knowledge encodements that synthesize research through Indigenous beadwork, textile, other fibre arts. Beading is Indigenous resurgence that connects me to my ancestors, and this research delves into what that means in a grounded wholistic way through my intersecting lenses of being a chronically ill neurodivergent Two-Spirit Mississauga Nishnaabe Lucbanin artist and scholar. Conceptualizations around the body, mind, spirit, land, and material expressions of culture are often thought of as separate entities due to colonization, so a foundational part of this work examines approaches to Indigenous ideas of wholeness in community and identifies what forms of decolonization and resurgence can facilitate reconnection with the spiritual. Beads come together and interplay with one another in similar ways that gained wisdoms do within the research process. While the overall design that is created through knowledge is powerful and important, so is every stitch that makes that design come into being. Each relative who collaborated on this dissertation brought a prismatic array of experiences and played a powerful role in shaping the trajectory of the ancestral knowledge encodement of this work in the Ngwaagan Regalia (2025). Throughout the dissertation are ancestral knowledge encodements—created through historical inspiration, depictions of relatives, tea-visits with kin, and narratives shared by family and community members. The encodements created and embedded into this written dissertation take the forms of photographs, historical images, digitally stitched collages, digital mixed media illustrations, paintings, and diagrams. I have chosen to honour this tradition of weaving in the threads of previous generations and connecting it to those in the future through integrating ancestral mkwendamwinan (memories) in the same way that I am including contemporary conversational dbaajmownan (stories).
Confidence in dynamic assurance cases
(2026) Diemert, Simon; Weber, Jens
Assuring safety- and security-critical systems is a necessary activity, both prior to deployment (at “design-time”) and during system operation (at “run-time”). The need for assurance is heightened as these systems increasingly depend on artificial intelligence and adaptation strategies to handle uncertainty in their operating environments. Assurance Cases (ACs) that incorporate structured arguments and supporting evidence are an important tool for establishing trust in critical systems. Modern ACs are not static documents: they are increasingly viewed as dynamic models of “through-life” assurance and are used for decision-making at both design- and run-time. These Dynamic Assurance Cases (DACs) incorporate dynamic sources of evidence and “live” data from development activities or operations (e.g., safety performance indicators). However, a question arises during their use: “are we confident in the claim(s) made by this version of the case?” While several methods exist to assess confidence in ACs, there is limited knowledge about their adoption by practitioners. Additionally, there are several limitations of quantitative methods, including: 1) an inability to consider the impact of dynamic aspects on confidence; 2) an inability to account for dialectic reasoning (i.e., “de-featers”); and 3) challenges related to subjectivity, interpretability, precision, and modelling nuanced reasoning. The overall objective of this work is to develop a new confidence assessment method that is grounded in the needs of practitioners and addresses the limitations mentioned above. Towards this objective, the main contribution of this dissertation is a new mixed (qualitative and quantitative) method for AC confidence assessment called Certus. With this method, confidence in an AC is modelled using vague, but linguistically meaningful, expressions (e.g., “I have very high belief that this claim is true”). A domain specific language is used to describe the propagation of belief through a structured argument to produce an overall belief valuation for the AC. Certus supports dialectic reasoning and can condition belief assessments on the availability of evidence and the value of performance indicators. The use of the language is guided by a methodology that integrates with the existing practices for developing (D)ACs. A denotational semantics for the language provides a formal basis for assessment. The language and method are evaluated through a series of analyses and a case study to demonstrate that they possess properties related to trustworthiness, including: propagation stability, sensitivity, expressivity, scalability, and applicability to DACs.
An innovative framework for next activity prediction using process entropy and dynamic attribute-wise transformer for business process monitoring
(2026) Zare, Hadi; Najjaran, Homayoun
In the field of Business Process Management (BPM), accurately predicting the next activity in an ongoing process is critical for improving operational efficiency, optimizing resource allocation, and enabling proactive decision-making. Although recent advances in machine learning (ML) and artificial intelligence (AI) have significantly improved predictive performance, several key challenges remain. These include: (i) the limited transparency of deep learning models when applied to datasets that may not require such complexity; (ii) the reliance on trial-and-error methods for selecting suitable models across diverse event logs; and (iii) the underutilization of attributes that can carry valuable information for prediction. Building on these identified gaps, this thesis is motivated by two central research questions: (1) How can predictive models be designed to effectively capture the complexity and variability inherent in modern event logs? and (2) How can organizations systematically determine the most suitable predictive models for their specific process characteristics? To address the first question, this thesis introduces a novel predictive architecture called the Dynamic Attribute-Wise Transformer (DAW-Transformer). The model enhances predictive capability by extending the standard transformer architecture through the integration of multi-head attention and a dynamic windowing mechanism tailored to each dataset. This design captures long-range dependencies across multiple event attributes, offering a richer and more detailed representation of process behavior and improving the model’s ability to generalize across heterogeneous logs. To address the second question, an Entropy-Driven Model Selection Framework is proposed. This framework employs process entropy as a quantitative indicator of event log complexity, enabling adaptive model selection that balances predictive accuracy and interpretability. By aligning model choice with dataset variability, it overcomes the limitations of existing approaches that apply a single predictive model indiscriminately across all process types, regardless of their structural diversity. The effectiveness of the proposed methods is validated through comprehensive experiments on six publicly available event logs across domains such as healthcare, logistics, and public administration. Results demonstrate that the DAW-Transformer achieves superior performance, particularly on high-entropy processes, where activities exhibit greater variability, while interpretable models such as Decision Trees perform competitively on low-entropy, structured processes. These findings highlight the value of aligning model complexity with process entropy and underscore entropy’s role as a guiding principle for model selection in predictive business process monitoring. In summary, this thesis contributes to: (1) improving interpretability and reducing trial-and-error model selection via an Entropy-Driven Model Selection Framework; (2) mitigating attribute underutilization through the DAW-Transformer; and (3) conducting comprehensive evaluations across diverse datasets. Together, these contributions enhance predictive accuracy, strengthen interpretability, and establish a data-driven foundation for adaptive model choice.
Reinforcement learning-based model predictive control for uncertain systems with application to robot manipulators
(2026) Lu, Tianxiang; Shi, Yang
Model-based control strategies often lead to performance degradation or even failure when implemented in practice since practical systems are subject to external disturbances and model uncertainties. Model predictive control (MPC) is also among these strategies. However, MPC is known to be capable of providing optimal control performance for constrained systems by solving an associated optimization problem. An essential component for constructing the MPC optimization problem is the prediction model which in many cases is the priori known physical model of the controlled system. To make MPC an effective method for practical systems with external disturbances and model uncertainties, adaptive MPC methods aim to obtain a satisfactory prediction model and thus control performance by performing system identification during the online control process. Despite the effectiveness of adaptive MPC, its intrinsic conservativeness leads to the development of other MPC frameworks that apply various techniques to provide robustness to the disturbed system with uncertainties. Reinforcement learning (RL) is one of the representative schemes that can be combined with MPC for safety guarantees and performance enhancement. In general, there exist three ways to combine RL and MPC. The first is to use RL to learn the optimal MPC parameters such as weighting matrices, constraint margins, and disturbance model parameters with the maximum long-term reward. The second applies MPC to check the safety of the actions generated by RL during the online control process. The last focuses on adopting MPC to generate the data for the offline training process of RL. In this dissertation, we mainly focus on the first way of combining RL and MPC to design novel reinforcement learning-based MPC (RLMPC) methods. Chapter 1 presents an overview of conventional MPC, adaptive learning-based MPC, RLMPC schemes along with classical control frameworks for robot manipulators. Chapter 2 introduces preliminary results on MPC implementation for linear and nonlinear systems as well as robot manipulators. Chapter 3 proposes a robust data-driven MPC for linear systems with mixed uncertainties via on-policy RL method with theoretical guarantees. Based on the proposed closed-loop MPC scheme, the on-policy SARSA algorithm is correspondingly chosen among various RL algorithms to construct a safe RLMPC framework with guaranteed recursive feasibility and closed-loop stability. Comprehensive comparative studies further demonstrate the advantages of better control performance and larger region of attraction over several well-established robust MPC and RLMPC methods. Building on Chapter 3, Chapter 4 develops an off-policy RLMPC approach for nonlinear systems with an integrated triggering mechanism. A quasi-dynamic event-triggered mechanism is incorporated into the proposed RLMPC framework to effectively reduce the overall computational load resulted from the parameter update by the off-policy RL. The resultant event-triggered RLMPC method is able to enhance the control performance while maintaining the closed-loop stability with guaranteed recursive feasibility and marginal conservativeness. Chapter 5 and Chapter 6 investigate the application of RLMPC schemes to robot manipulator with relatively simple structures and complex multi-joint structures, respectively. In Chapter 5, a variant of RLMPC methods proposed in Chapter 3 is further developed for manipulators with accessible dynamic model however subject to both model uncertainties and external disturbances. Therefore, the on-policy SARSA RL algorithm is again utilized to construct the policy for updating the uncertain Coriolis and centrifugal matrix, as well as the potentially imperfect feedback control gain. A case study on a planar robot manipulator with two revolute joints shows the capability of achieving better closed-loop performance under the proposed method. Chapter 6 presents the experimental results of applying a RLMPC scheme to the trajectory tracking problem of a practical robot manipulator: UR-10e. Considering the ability to provide real-time control actions, the trajectory tracking task is converted to the consecutive waypoint reaching task to further facilitate the design of a lightweight RLMPC framework. Two trajectories are adopted in the experimental validation to demonstrate superior performance under the proposed method over conventional MPC and classical soft actor-critic method, especially in the presence of relatively large model mismatch. Lastly, Chapter 7 concludes this dissertation and introduces potential future research directions.
Scripts and screens: A cross-cultural analysis of gendered dating advice in short-form video content on Instagram and Red
(2026) Huang, Jindi; Garlick, Steve
This thesis examines how short-form dating advice videos shape contemporary understandings of dating, relationships, and romantic behaviour on social media. Focusing on Instagram Reels and the Chinese platform Xiaohongshu (Red), the study explores how dating content creators construct and circulate knowledge about heterosexual relationships in digital environments. Using a qualitative comparative approach, the research analyzes twenty short videos by four creators across Instagram and Red. The analysis combines thematic analysis and critical discourse analysis, while drawing on Foucault’s concepts of power, knowledge, and discourse, Simon and Gagnon’s sexual script theory, and feminist perspectives. The findings identify three key themes: gender roles and dating scripts, strategy-based advice and discourse patterns, and cross-cultural comparisons. While many short videos frame the dating advice as empowering, traditional heterosexual scripts continue to shape expectations around women’s emotional labour and relational behaviour.
Modelling nanocomposite dielectrics - polarization and inclusions at the atomic scale
(2026) Adluri, Archita; Paci, Irina
This thesis details the computational investigations into molecular-nanoparticle based nanocomposite dielectric candidate materials. Mainly silver-based nanoparticles under 32 atoms in metal-oxide and polymer model materials to understand how encapsulation affects local and bulk dielectric function and how to model such candidates. Chapter 1 introduces the ideas behind molecular nanoparticles (MNP) and functional dielectrics at the sub-100nm range. Emphasis on bridging gaps present in the literature at present and potential novel uses between atomic-doping and traditional larger nanoparticles. Chapter 2 delves into theory and background as this is an entirely theoretical and computational thesis. An explanation of both static and optical dielectric starting from electronic structure and ending in a comparison of methods. Further computational details provided in relevant next chapters. Chapter 3 contains details of the major topic of this thesis: How to model a molecular nanoparticle inclusion in a solid-state dielectric. The candidates of Ag inclusion and MgO candidate metal-oxide were modelling using Car-Parinello Molecular Dynamics to obtain a size and shape correlation in nanoparticles as well as some guidelines for computing charge distribution and quantum of polarization effects. Chapter 4 contains details of the optical polarization of Ag-PVDF nanocomposites with a focus on interacting particles and non-homogeneity in MNP distribution. This project aimed to move towards a more realistic model with two nanoparticles interacting at various distances. Chapter 5 is an attempt to push the boundaries of highly detailed molecule-level polarization calculations and a discussion on their usefulness without experimental backing. Chapter 6 summarizes this thesis followed by some QuantumEspresso code and parameters used.
Gardens on the edge: Estuarine root gardens as places of tangible heritage and Indigenous futurity
(2026) Maurice-Hammond, Isabelle K.; Mathews, Darcy
Estuarine root gardens are Indigenous plant stewardship landscapes on the Pacific northwest coast of North America. The plants that were principally cared for in these coastal sites–Pacific silverweed (Argentina egedii (Wormsk.) Rydb.; syn. Potentilla pacifica (L.) Howell), springbank clover (Trifolium wormskioldii Lehm) and northern rice-root lily, Fritillaria camschatcensis (L.) Ker Gawl)– were eaten by Indigenous Peoples from northern California to Alaska. In certain areas, generations of Indigenous Peoples cared for and altered the soils, hydrology, and ecology of high estuarine marshes to maximize the growth and output of these nutritionally, spiritually, and economically important plants. Though stewardship methods and the location of some of these places are still known by Indigenous Knowledge Holders, descendant communities, and researchers (primarily in coastal British Columbia, Canada), the extended dismissal of Indigenous plant stewardship practices by settler ethnographers and archaeologists has resulted in a chronic under-representation of these places in the archaeological record of British Columbia. Furthermore, tangible remnants of Indigenous estuarine stewardship–in the form of legacy ecosystems and cultural soils–is not currently seen as sufficient evidence in categorizing these places as archaeological sites. This has implications for their protection under current provincial Heritage and Conservation (HCA) legislation, as well as the ability of descendant communities to reconnect with these culturally significant places and foods. Finally, identifying estuarine root gardens, and contributing to eco-cultural restoration efforts, is occurring within a context of ongoing and cumulative colonial violence to Indigenous territories, with estuaries particularly at risk. Understanding and addressing these complex factors is key to the successful restoration, renewal, and creation of stewarded places. Ultimately, this dissertation aims to address the disconnect between the (often limited and fragmented) archaeological study of estuarine root gardens and the importance of these landscapes for Indigenous heritage, food sovereignty and security, and cultural reconnection through the Indigenous-led restoration and renewal of food systems. In doing so, I present and tests a series of novel and interdisciplinary methods to identify estuarine root garden sites and better understand their post-stewardship trajectories, accounting for eco-cultural context, present day ecologies and hydrologies, and cumulative colonial impacts. Combining archaeological excavations, ecological monitoring, and pedological analysis, I examine two estuarine root gardens (and comparative periphery, or control sites) occurring in different eco-cultural contexts. The first, at Tl’chés, is a Lək̓ʷəŋən/Songhees root garden that was no longer known by descent communities and knowledge holders, making it the first of its kind to be formally identified in Coast Salish territories. The second, at the mouth of the Gwa’ni (Nimpkish) River, is a known ‘Na̲mg̲is Kwakwaka'wakw estuarine root garden site which has been documented by non-Indigenous researchers since the 19th century, offering a rare glimpse at more than a century of changing use and occupation patterns in the area. A third estuarine root garden at Tsinwilht’as (Anderson) Creek in ʕaaḥuusʔatḥ Nuu-chah-nulth territories was investigated as part of an Archaeological Impact Assessment (AIA) by Stafford (2020). As such, the methods described in this paper could not be replicated there. However, the thoroughness of Stafford’s (2020) investigation provides important comparative data, helping illustrate both the similarities and differences of estuarine root gardens located in different eco-cultural areas, as well as an example of best archaeological practice in the field of Cultural Resource Management (CRM). Overall, the methods applied during this research demonstrated that 1) estuarine root gardens continue to support distinct ecological assemblages shaped by past Indigenous stewardship, though the ecological trajectory of these places vary based on local hydrological, cultural, and climactic factors. 2) Estuarine root garden soils are morphologically and chemically distinct, with heightened levels of organic matter and available soil phosphorous (P). As such, they can best be understood as cultural or Indigenous soils, developed in situ to increase garden productivity and access to preferred long, straight roots and rhizomes. 3) Archaeological signatures at estuarine root gardens are variable; as such, archaeologists need to expand their arsenal when investigating a potential estuarine root garden site. These considerations, furthermore, should be rooted in engagement with the communities to which these places belong. 4) Finally, estuarine root gardens are living archaeological sites. As such, they are places with profound ties to the communities that built and cared for them, a relationship that extends into the future. Developing a better understanding of estuarine root garden formation has implications for the community-driven restoration of these places, often occurring in contexts where coastlines have been altered by ongoing and cumulative settler colonial impacts.
GPU-based lock-free mesh reduction using deterministic vertex clustering
(2026) Hanif, Khizra; Chester , Sean
The growing demand for real-time rendering and processing of complex 3D models in film production, gaming, and scientific visualization has exposed the limitations of CPU-based mesh simplification techniques. While S-Weld ensures deterministic clustering, its sequential execution makes it unsuitable for large-scale meshes. The multi-core P-Weld improves performance through lock-free multi-threading but remains constrained by CPU core count and memory bandwidth. To address these issues, this thesis presents a GPU-accelerated vertex clustering framework that extends the deterministic behavior of P-Weld to a fully parallel and memory-adaptive GPU architecture. The work begins with a direct CUDA port of P-Weld and introduces an On-the-fly neighbor evaluation method that performs clustering without storing explicit adjacency lists, followed by a fully GPU-resident sparse voxel-grid framework for efficient processing of large meshes. Shared-memory caching, warp-synchronous centroid updates, and sparse neighbor filtering are used to minimize redundant computations and improve parallel efficiency. Existing GPU-based libraries, such as FRNN and cuNSearch, were evaluated but found unsuitable for large irregular meshes on consumer laptops. A custom sparse grid-based neighbor search was therefore developed to perform efficient ε-neighborhood queries entirely on the GPU within limited VRAM. The proposed pipeline was evaluated on five benchmark meshes, including Bunny, Lucy, Thai Statue, Manuscript, and the point cloud LiDAR dataset using various clustering thresholds. The GPU versions achieved identical results while providing a 10-26× speedup, improving the scalability of vertex clustering for large 3D mesh simplification. In summary, this thesis presents the first deterministic GPU extension of vertex clustering, introducing a memory-efficient sparse-grid neighbor search and a fully parallel pipeline that reproduces accurate results, advancing scalable and reproducible mesh simplification.

Browse

Recent Submissions