26734 research outputs found
Sort by
Enhancing Leak Detection in Oil and Gas Pipelines with Multi-Task LSTM Model
No full textOil and gas pipelines form a critical component of Canada’s industrial infrastructure, providing an efficient means of transporting natural resources across vast geographic regions. However, leak events pose severe environmental and economic risks, reinforcing the need for robust, accurate, and explainable monitoring systems. This thesis proposes a compact, multi-task Long Short Term Memory (LSTM) model for joint leak detection and localization in multivariate time-series data, leveraging the model’s ability to capture long range temporal dependencies.
All datasets used for training and evaluation are generated using a Real Time Transient Model (RTTM), which simulates the fluid dynamic response of eight pipeline zones under leak and no-leak scenarios. The simulations incorporate conservation equations and the GERG-2008 equation of state to model pressure, flow, and temperature dynamics during leak propagation. Each sample encodes a sequence of sensor readings along the pipeline, with labels identifying leak presence and leak position. The LSTM model is evaluated under consistent 50/25/25 train–validation–test splits and benchmarked against a conventional one dimensional CNN baseline to ensure fair architectural comparison. Performance is assessed using classification metrics (accuracy, precision, recall, F1-score, and ROC–AUC) as well as localization metrics (MAE and Hit@5). Across all eight pipeline zones, the LSTM achieves near perfect classification performance and produces more accurate leak localization than the CNN, particularly in zones with longer sequences or widely spaced sensors. Latency and throughput measurements further demonstrate that the LSTM remains computationally feasible for real-time deployment. An ablation study evaluates the impact of architectural variations such as hidden size, number of layers, and activation functions revealing that the final LSTM configuration offers the best balance between generalization, localization accuracy, and inference speed. The localization output also contributes to model explainability by narrowing the region of interest for operators, aligning with the needs of real-world decision support environments. The proposed approach integrates with the PRIME and SHAPE research platforms, supporting operator facing visualization and training workflows. Limitations regarding reliance on simulated data, lack of cross validation, and assumptions about sensor accuracy are discussed, alongside recommendations for future work involving interpretability tools, distance aware loss functions, and testing with live or replayed SCADA datasets. Overall, this research demonstrates that sequence-based modeling of pressure flow dynamics can substantially enhance the reliability, interpretability, and operational readiness of pipeline leak detection and localization systems, contributing to safer and more sustainable energy infrastructure
Maternal Inflammation as a Mediator of Associations Between Pre-Pregnancy Diet, and Brain Connectivity and Cognitive Outcomes in Young Children
No full textMaternal diet, even prior to conception, supplies critical nutrients that are foundational for fetal brain development. Maternal diet has also been associated with BMI and inflammation. Pre-pregnancy BMI and prenatal inflammation have been linked to child cognitive and brain development. Thus, diet prior to pregnancy, along with pre-pregnancy BMI and maternal inflammation, may be associated with child intelligence, executive function (EF), and brain network connectivity. However, no research to date has examined the associations between maternal pre-pregnancy diet and child structural brain networks, intelligence and EF. Furthermore, the mediating effect of maternal inflammation on this association has not been studied. This thesis aimed to fill this gap in knowledge, examining the associations between pre-pregnancy diet and child salience network (SN) and default mode network (DMN) structural connectivity as well as intelligence and EF. The mediating effect of maternal prenatal plasma C-reactive protein (CRP) on all the above associations was also investigated. This thesis used existing data from the Alberta Pregnancy Outcomes and Nutrition (APrON) cohort and the Calgary Preschool Magnetic Resonance Imaging (MRI) Dataset. For both studies, maternal diet the year prior to pregnancy was quantified using the Dietary Inflammatory Index (DII) and Mediterranean Diet Adherence (MED). Both studies used maternal plasma samples collected in the third trimester to quantify CRP levels. Covariate-adjusted, pre-pregnancy BMI-stratified, robust regressions examined associations between DII and MED scores, and brain metrics as well as intelligence and EF. Mediation analyses examined if maternal plasma CRP mediated these associations. Sex-stratified associations were explored. In the first study, I used data from children who underwent a diffusion tensor imaging (DTI) scan between the ages of 3 – 7 years. Structural connectomes for the SN and DMN were reconstructed and metrics were calculated using graph theory analysis. In women with a lower pre-pregnancy BMI, higher maternal MED scores were associated with greater DMN global and local efficiency, though this relationship only survived in females. No significant mediation effects were noted. In the second study, I used data from children who completed the Wechsler Preschool and Primary Scale of Intelligence – IV Canadian (WPPSI-IVCND) and the Behaviour Rating Inventory of Executive Function – Preschool (BRIEF-P) between ages 3 – 5 years, to assess intelligence and EFs. In women with higher pre-pregnancy BMI, an anti-inflammatory pre-pregnancy diet (i.e., lower DII and higher MED) was associated with higher WPPSI-IVCND Verbal Comprehension Index (VCI) scores in children. Sex-specific associations were found but did not survive correction for multiple comparisons. No significant mediation effects were noted. Taken together, these studies found novel evidence of relationships between pre-pregnancy diet and child DMN structural connectivity and intelligence, which differed based on maternal pre-pregnancy BMI. These results provide initial support for the creation of community knowledge dissemination efforts promoting proper nutrition in women who are planning to conceive, aiming to support child brain and cognitive development
Playing at Technological Boundaries: Participatory Art on Algorithmic Decision-Making
No full textToday, machine learning and predictive algorithms find patterns in large data stores and make predictions which corporations and governments use to support decision making. These systems’ representation of reality can be more important to decision-making than the complex reality they are intended to reflect. Such situations become problematic when they undermine the responsiveness and inclusivity of public decision making and when their use creates, introduces, or perpetuates social or economic inequality. To address these challenges, the general public must be able to participate in discourse about the social implications of the use of algorithmic decision-making. My research investigates how participatory art experiences might provide a context for the public to consider the impact of algorithmic decision-making on society. I created a series of participatory art installations concerning the impacts of algorithmic decision-making, providing contexts for enactive exploration of such systems. The installations were: Algorithmic Rituals, using creation to reflect on how technology habits impact interactions with others; Entanglements using narrative to contemplate how the impacts of algorithmic systems can intersect; and The Neural Network, using play to construct a collaborative neural network. I conducted phenomenographic interviews to better understand how attendees experience art installations on technical topics. The data provided insight for the installations as I developed them, and revealed how participants encountered ideas and content outside of insight for other researchers looking to support public consideration of new and challenging ideas about technology. I argue that experiencing discomfort at the boundaries of attendees’ expertise is a vital aspect of supporting the emotional capacity to respond to an evolving technological and social landscape. their expertise in the works. They pivoted from being unsure about the experience to open to the ideas they encountered. In this thesis I discuss these pivots as instances of boundary play, and provide insight for other researchers looking to support public consideration of new and challenging ideas about technology. I argue that experiencing discomfort at the boundaries of attendees’ expertise is a vital aspect of supporting the emotional capacity to respond to an evolving technological and social landscape
Expanding our understanding of adult-onset myopathy: contributions of satellite cell dysfunction and denervation
No full textAdult-onset genetic myopathies are amongst the most challenging clinical cases to diagnose. These conditions are characterized by extensive genetic heterogeneity, overlapping clinical phenotypes, and poor genotype-phenotype correlation. Symptom onset can occur at any age, and is often confused for acquired disease. As a result, patients often experience a lengthy diagnostic odyssey, in which they can spend years or even decades with diagnostic uncertainty, and without clear guidance regarding the risk to their family members. These conditions are nearly all untreatable due to our lack of knowledge of the underlying disease mechanisms. Genetic myopathies arise from variants in many different proteins. Symptoms begin with selective muscle involvement and gradually progress to include most skeletal muscles, and in some cases, cardiac or respiratory muscles. Prior research has linked some myopathies with differentiation defects and changes at the neuromuscular junction (NMJ). Our work aimed to uncover shared biological disease mechanisms among adult-onset genetic myopathies. We analyzed skeletal muscle biopsies from 10 individuals with genetically confirmed myopathies and 4 unaffected controls. Single nucleus RNA sequencing revealed a group of myonuclei enriched in myopathy that co-express markers of mature muscle and NMJ-associated genes. These nuclei also express key indicators of a denervation response, including RUNX1, CHRNG, NCAM1, and MYOG. Their abundance correlated with the severity of muscle pathology, independent of age or sex. Spatial transcriptomics localized these nuclei to small, angular fibers consistent with denervation and immunostaining showed disrupted patterns of innervation. Analysis of public datasets confirmed elevated expression of NMJ-related genes in other myopathies, notably myotonic dystrophy and polymyositis. These findings suggest that adult myopathic muscle activates a denervation response that may play a role in disease progression. The consistent presence of similar transcriptional signatures across multiple myopathies indicates that denervation may be a common feature of these conditions
Improving Precise Smartphone GNSS with Robust Dynamics, Adaptive Stochastics, and Cycle Slip Repair
No full textThe proliferation of Global Navigation Satellite System (GNSS) chipsets in mass-market smartphones has democratized access to positioning data, however, attaining decimeter-level precision remains a challenge due to the limitations of low-cost hardware. Smartphone observations are characterized by low Carrier-to-Noise density (/!), high susceptibility to multipath, and frequent loss of phase lock. Standard positioning algorithms, originally designed for geodetic-grade receivers, struggle in this environment as they rely on heuristic dynamic models, static stochastic weighting, and rigid integer ambiguity resolution strategies that discard corrupted phase data. This thesis proposes a comprehensive framework to enable precise smartphone positioning by addressing three critical failure points: erratic user dynamics, environmental volatility, and carrier-phase discontinuity. First, a Robust Dynamics model is developed using a Doppler-Based Prediction (DBP) technique. By using Doppler-derived velocity for state propagation, this method automates the estimation of process noise (Q), allowing the filter to instantaneously adapt to unconstrained user motion. Experimental validation demonstrates that DBP reduces horizontal positioning errors by approximately 57% in kinematic scenarios compared to standard models. Second, an Adaptive Stochastic model is implemented using Variance Component Estimation (VCE). This Adaptive Kalman Filter (AKF) learns the true measurement noise (R) in real-time, effectively down-weighting multipath-affected signals without discarding them. This approach yields horizontal accuracy improvements of 35% in static and 27% in vehicular environments compared to traditional elevation and /! based weighting. Finally, the research introduces a hierarchical Cycle Slip Detection and Repair (CSDR) framework. Moving beyond the standard "detect-and-reset" paradigm, which destroys filter convergence, this method employs a hybrid detection scheme and a stochastic repair engine. Using Partial Ambiguity Resolution (PAR) with a Fixed Failure-Rate Ratio Test (FF-RT), cycle slips are estimated as integers and restored into the filter with associated variance. Validation on Google Pixel 4 datasets confirms that this stochastic repair strategy maintains phase continuity, reducing vertical positioning RMSE by approximately 45% compared to standard reset methods. Collectively, these contributions demonstrate that high-precision smartphone positioning is achievable not by filtering out noisy data, but by rigorously modeling its dynamic and stochastic characteristics
Synthesis and thermo-oxidative kinetic analysis of cellulose microfibers from palm leaves using ammonia fiber expansion
No full textAbstract Global urbanization is driving high volumes of agricultural and food waste, creating an urgent need for sustainable and effective technologies to convert biomass into valuable products. This study explores the conversion of palm waste into cellulose microfibers (CMF) using Ammonia Fiber Expansion (AFEX) followed by acid hydrolysis, with a focus on structural characterization, thermal stability, and reaction kinetics compared to raw material. The resulting CMF exhibited elongated, uniform fibers with smooth surfaces, with lengths of 0.1–3.0 mm, and diameters of 5–20 μm. X-ray analysis revealed a significant increase in the carbon/oxygen ratio, from 1.8 ± 0.2 in raw palm leaves to 2.7 ± 0.3 in CMF, indicating enhanced carbon content due to dehydration and reduction of carbonyl groups. FTIR spectra confirmed effective removal of lignin and hemicellulose after treatment, further supporting this chemical transformation. Thermal analysis demonstrated that CMF possesses higher heat content than raw leaves, suggesting its potential for energy-related applications. TGA showed that CMF decomposes at slightly higher temperatures, indicating improved thermal stability. Isoconversional kinetic analysis using the Vyazovkin Nonlinear (NLN) and Kissinger-Akahira-Sunose (KAS) methods revealed variable effective activation energies (Eα), consistent with a complex degradation mechanism. Overall, CMF displayed lower Eα values than raw biomass, especially at early and mid-reaction stages. Kinetic modeling at 50% conversion showed a markedly higher pre-exponential factor (Aα) for raw leaves (2.8 × 10¹³ s⁻¹) compared to CMF (7.4 × 10⁹ s⁻¹), reflecting structural alterations from treatment. Both raw and CMF samples exhibited negative activation entropy (ΔS≠) values of − 237.7 and − 240.3 J mol⁻¹ K⁻¹, respectively, suggesting greater molecular order in activated complexes. The enthalpy of activation (ΔH≠) was 149.7 ± 3.9 kJ mol⁻¹ for raw leaves versus 120.4 ± 3.9 kJ mol⁻¹ for CMF, Gibbs free energy of activation (ΔG≠) was slightly higher for raw leaves (297.0 ± 3.9 kJ mol⁻¹) compared to CMF (269.4 ± 3.9 kJ mol⁻¹), primarily due to differences in ΔH≠. These kinetic parameters are crucial for any future implementation of palm leaves conversion into CMF at the industrial scale
The Theoretical and Practical Implications of Entrepreneurial Ecosystem Orchestration
No full textThis dissertation advances understanding of entrepreneurial ecosystem orchestration—the deliberate facilitation of interactions and actions among independent ecosystem actors. It strengthens the underlying assumption of interdependence of the entrepreneurial ecosystem (EE) concept, positing that purposeful orchestration is essential for cultivating and strengthening EE interdependence, which in turn enhances the ecosystem-level outcome of productive entrepreneurship. The first study examines how entrepreneurs’ engagement within EEs can be purposefully facilitated through advisor-led ecosystem orchestration. It conceptualizes ecosystem orchestration as a problem-oriented process in which advisors strategically facilitate entrepreneurs’ interactions within the ecosystem to enhance their problem-solving effectiveness. Ecosystem advisors are foregrounded as key orchestrators who leverage externally oriented dynamic capabilities to support problem-oriented orchestration. Using longitudinal qualitative data from 30 entrepreneurs, the study develops a problem-oriented orchestration model that identifies problem identification (sensing), engagement facilitation (seizing), and advice provision (reconfiguring) as core orchestration actions performed by ecosystem advisors and links these actions to entrepreneur-level antecedents and outcomes. The second study extends the problem-oriented orchestration approach by incorporating an intersectional lens. It explores how problem-oriented orchestration enables underrepresented entrepreneurs with intersecting identities to address business challenges. Employing a multiple-case study method with an embedded design, the study includes 15 entrepreneurs stratified by gender and immigration status, and 44 business problems they identified. This research design enables the examination of how the problem-oriented orchestration approach generates differentiated impacts on entrepreneurs with intersecting identities, arising from disparities in their human and social capital. The findings reveal that the intersection of entrepreneurs’ gender and immigrant status shapes distinct configurations of human and social capital, rendering problem-oriented ecosystem orchestration a compensatory, complementary, lagged, or supplementary role in facilitating entrepreneurial problem-solving. The third study broadens the scope of ecosystem orchestration to a collective form, in which multiple actors jointly shape goals, actions, and interactions. Situating collective orchestration within the context of student development, the study integrates student development theories with ecosystem orchestration research to explain how universities, industry partners, and students undertake interdependent yet complementary orchestration roles. Drawing on a single-case study of a university training program, the study introduces a multi-level collective orchestration framework: program administrators act as institutional orchestrators designing program and governance structures; industry partners serve as organizational orchestrators fostering students’ relational integration; and students operate as individual orchestrators managing behavioural engagement across institutional boundaries. Together, these three studies advance ecosystem orchestration research by explaining how ecosystem orchestration operates across diverse contexts to promote productive entrepreneurship through enhancing problem-solving, supporting underrepresented entrepreneurs, and enabling collective talent development
Patient safety and Errors in Veterinary Anesthesia
No full textVeterinary anesthesia is a specialty in which medical errors can lead to significant patient harm. The aim of this thesis was to identify factors contributing to errors in veterinary anesthesia, evaluate the effectiveness of a pre-induction checklist in reducing omission errors, and develop a quality improvement initiative. Four studies were conducted and a Veterinary Anesthesia Error Reporting System was developed. The first study was a literature review on medication errors in veterinary anesthesia. The second study collected voluntarily reported perianesthetic medication errors in small animals. The third and fourth studies were conducted during a live animal teaching session with veterinary students at the University of Calgary’s Faculty of Veterinary Medicine (UCVM). The third study evaluated dose calculation errors, while the fourth study assessed the effectiveness of an anesthesia safety checklist in reducing incomplete safety tasks. Medication errors were found to be the most commonly reported error in veterinary medicine. A medication error rate of 1.8% (48 errors in 2,728 anesthesia/sedation procedures) was identified in veterinary clinics in Calgary. Most reports (69% [33/48]) were near misses (an error that did not reach the patient). Dose calculation errors were the most common type reported; 50% (24/48). During a live animal teaching laboratory session, in which 83 anesthesia procedures were performed, a dose calculation error was identified in 10.8% (9/83) of anesthesia protocols. Furthermore, the use of a pre-induction checklist identified at least 1 omitted pre-induction task in 67.5% (56/83) of anesthesia procedures. The most frequently missed item was premeasuring the endotracheal tube insertion depth (42.2%, 35/83) and the checklist identified closed adjustable pressure-limiting valves in 4.8% (4/83) of cases. A reporting system specific for veterinary anesthesia was developed with the features recommended by the World Health Organization. In conclusion, the results of these studies identified several factors contributing to errors in veterinary anesthesia, with dose calculation errors as major contributor. The results support dose independent double-checking, anesthesia chart reviews, and the use of checklists. Finally, the developed Reporting System received endorsement from an international veterinary anesthesia organisation and shows strong potential as a successful quality improvement initiative
Design, Synthesis and Self-assembly of Amphiphilic Cyclodextrin-Based Materials for Enhanced Ion Conduction and Drug Encapsulation
No full textSupramolecular chemistry, often referred to as ‘chemistry beyond the molecule,’ explores the intricate world of non-covalent interactions that govern the assembly, recognition, and function of complex chemical systems. Supramolecular systems enable the design of highly functional materials with applications spanning many fields including drug delivery, catalysis, and nanotechnology. This thesis explores the synthesis, characterization, and potential applications of several new families of cyclodextrin (CD) derivatives. These novel derivatives were designed to investigate their supramolecular assembly, aiming to develop innovative alternatives in energy applications. To prepare readers with the required background to understand the topics of this thesis, Chapter 1 is dedicated to introducing the fundamental aspects of CDs, crown ethers, liquid crystal materials and their associated mesophases commonly formed in solid states as well as characterization methods. A brief review on amphiphilic CD-based liquid crystalline materials during recent years has also been provided. Following the introduction, Chapter 2 details the synthesis of family of amphiphilic β-CD derivatives. This chapter presents an effective strategy for introducing 12-benzocrown-4 and 15-benzocrown-5 functionalities at the termini of seven oligoethylene glycol (OEG) chains, which are conjugated to the primary face of the β-CD scaffold. In combination with fourteen stearoyl chains on the secondary face, these new amphiphilic CD derivatives have shown unprecedented ability to form bicontinuous cubic mesophases. Chapter 3 builds upon the findings of the previous chapter by enhancing the synthetic efficiency and investigating the
‘open’ crown-ether concept, achieved by incorporating linear OEG chains onto a benzene ring in place of a closed crown structure. To capitalize on the potential for enhanced lithium-ion mobility within the 3D bicontinuous cubic mesophase, these molecules are further evaluated for lithium conduction, confirming their superior transport properties compared to previous systems. Chapter 4 leverages epichlorohydrin chemistries to develop a novel family of amphiphilic β-CD-based liquid crystalline materials that are polyesterified with 14 varying lengths of aliphatic chains at the secondary face and 14 oligoethylene glycol (OEG) chains at the primary face, likely approaching a nearly ‘cylindrical’ geometry. By varying the lengths of the
aliphatic chains, we present a more systematic investigation in which the lengths of the hydrophobic chains were methodically varied to produce a series of amphiphilic β-CD derivatives with different hydrophilic-to-hydrophobic volume ratios. This systematic approach enabled the identification of an optimal ratio range for the formation of bicontinuous cubic mesophases. Chapter 5 focuses on the molecules that did not form liquid crystalline materials from the previous chapter. Instead, they were tested for micelles and their effectiveness for potential application in host-guest inclusion studies with APIs (active pharmaceutical ingredients). Overall, this work advances CD research by addressing synthetic challenges, expanding mesophase design principles, and explore applications in ion conduction and molecular encapsulation
Dynamics of Colloids at Equilibrium and Thermal Non-equilibrium: From Microrheology to Environmental Sensing
No full textColloidal dispersions display diverse dynamics depending on whether they are at equilibrium or driven out of equilibrium. At equilibrium, colloids undergo Brownian motion and govern processes such as self-assembly and phase transition. When exposed to thermal gradients, colloids exhibit driven motion that is useful in separation, enrichment, and biochemical sensing applications. This motion is sensitive to surface chemistry, system composition, and background temperature. In the first part of this thesis, colloid dynamics at equilibrium are exploited for microrheological characteri-zation of polymer solutions in different polymer concentration regimes. For this end, generalized theoretical equations are developed for the mean squared displacement and specified for different rheological models. We demonstrate that the polymer concentration regimes can be distinguished using the fractional rheological parameters. We further propose simple approximations for the critical overlap concentration and the shear viscosity of viscoelastic liquidlike solutions. At thermal non-equilibrium, we examined the thermophoretic transport of colloid particles in different liq-uid media. In non-polar polymer solutions, we extracted the van der Waals (vdW) interactions from ther-mophoresis measurements. This was achieved by developing a theoretical framework for colloid thermophore-sis in polymer solutions. The theory reveals the influence of vdW interactions and polymer concentration on colloidal thermophoresis. A non-monotonic dependence of colloid thermophoresis on polymer concentration was observed and attributed to the opposing effects of increased polymer concentration and higher solution viscosity on polymer distribution. In aqueous solutions, the sensitivity of thermophoretic motion to surface chemistry was exploited to detect per- and polyfluoroalkyl substances (PFAS) adsorbed on various model microplastics. It was observed that the adsorption of PFAS molecules on particles produces distinct thermophoretic responses according to the chain length and head group of PFAS. Through the application of the mode-coupling model (MCM) for thermophoresis, we correlated the number of adsorbed PFAS molecules with the number of water molecules in the hydration shell around the colloid. This work addresses how colloidal dynamics can be utilized at equilibrium for soft matter characterization and thermal non-equilibrium for measuring intermolecular interactions and sensing toxic chemicals. The findings highlight strategies for the development of colloid-based microfluidic PFAS sensors and controlling particle motion in polymer solutions for colloidal printing and particle detection applications