26734 research outputs found
Sort by
Real-Time Bus Arrival Time Prediction: A Practical Data-Driven Approach
No full textWith growing support for diversifying travel options and efforts to improve sustainability in the transportation industry, transit system growth is anticipated worldwide. With this, comes an opportunity to leverage new vehicle tracking technologies and big data to improve transit services and attract users. The demand for reliable transit systems paired with widespread access to transit data highlights a need for a model capable of accurately predicting transit travel time. Although numerous studies have explored the bus travel time prediction modelling problem, thus far no method has emerged as consistently superior, and most complex models are difficult to interpret. Previous research has focused primarily on short-term predictions and few studies have investigated the effects of policy adherence and driver behaviour on prediction accuracy. The goal of this thesis is to develop a practical data-driven model capable of accurate bus arrival time prediction that can incorporate real-time data and generate predictions for a medium-term forecast horizon, capture real-world transit conditions, and provide a high level of transparency and interpretability. Following extensive data compilation and consolidation, including the development of a novel process for combining APC and AVL data, a bus arrival time prediction model is developed. Bus arrival time is predicted at downstream stops on the same trip or the upcoming trip by combining link travel time and stop dwell time components. Each component is predicted using historical observations and real-time information while considering operational challenges and strategies. The model performs well, particularly for a medium-term forecast range. It outperforms scheduled predictions for forecast horizons exceeding 30 minutes and has a mean absolute error (MAE) below six minutes for forecast horizons up to two hours. Prediction error is generally proportional to the forecast horizon and any problematic route segments or stops are easily identifiable. Overall, the model’s transparency, interpretability, and prediction accuracy make it a practical solution for bus travel time forecasting. It enables transit agencies to identify system weaknesses and improve service reliability by adjusting services or operations and provide more accurate information to passengers. Ultimately attracting users, reducing traffic congestion, and improving the safety and sustainability of the transportation network
Healthcare Cost Associated with an Acute Mental Healthcare Bundle in Pediatric Emergency Departments
No full textAbstract Objective To evaluate healthcare cost associated with a standardized mental healthcare bundle in two tertiary pediatric emergency departments (EDs) in Alberta, Canada. Methods We linked data from the two EDs and administrative health databases. Patients < 18 years presenting with mental health concerns were enrolled during two periods: pre-implementation (standard care) and implementation (mental health bundle). Healthcare resource utilization (HCRU) and costs were compared over 12 months following the initial ED visit. Results The study included 686 pre-implementation (mean age 12.6 years) and 692 implementation (mean age 13.0 years) patients with similar sex distribution (p = 0.65). Implementation patients had lower admission rates at the index ED visit (4.6% versus 9.8%; p < 0.001). During 1 year of follow-up, implementation patients had fewer hospitalizations (mean = 0.4 versus 0.5; p < 0.001), non-ED ambulatory care visits (mean = 2.7 versus 10.0; p < 0.001), and practitioner claims (mean = 34.3 versus 46.8; p = 0.001). Implementation patients had lower healthcare costs (10,336–19,326 [95% CI 22,056]; p < 0.001) attributable to lower hospitalization (5257–9994 [95% CI 11,782]; p = 0.01), ambulatory care (1383–4027 [95% CI 4727]; p < 0.001), and practitioner (3410–5306 [95% CI 5935]; p < 0.001) expenses during the follow-up. After risk adjustments, care bundle was associated with a 40% [95% CI 26–52%] reduction in mental health-related healthcare costs (p < 0.001) and 37% (95% CI 23–49%) reduction in all-cause healthcare costs (p < 0.001). Conclusions Integrating a standardized mental healthcare bundle in pediatric EDs was associated with less healthcare use and cost savings. These findings support its broader use to enhance mental healthcare delivery in emergency settings
Rivers With Borders: Bi-National Water Management Model for the St. Mary-Milk Basins
No full textTransboundary water management in semi-arid basins is increasingly challenged by climate variability, aging infrastructure, and rigid allocation frameworks. The St. Mary and Milk River Basin, shared by Canada and the United States, is governed by the 1909 Boundary Waters Treaty and the 1921 International Joint Commission Order, which allocate water based on reconstructed natural flow. Despite this framework, both countries routinely experience unmet entitlements due to hydrologic variability, operational constraints, and limitations in diversion and storage infrastructure. These challenges have gained urgency following recent droughts and infrastructure failures that have exposed vulnerabilities in the existing system.
This thesis applies a bi-national water management model of the St. Mary and Milk River Basin using the Water Resources Management Model. A detailed Base Case configuration is constructed to replicate current infrastructure, operating rules, and naturalized inflows over the 1982 to 2014 period. The model is then used to evaluate structural and administrative scenarios designed to reduce entitlement shortfalls. Structural scenarios include increasing the diversion capacity of the St. Mary Canal from 600 to 850 cubic feet per second and the addition of an on-stream storage reservoir at the Forks location in the Canadian Milk River basin. Administrative scenarios examine alternative credit accounting frameworks that modify how surplus and deficit deliveries are tracked within a water year. Results show that increasing canal capacity improves United States entitlement receipt on the St. Mary River but does not eliminate shortfalls due to the timing and short duration of peak snowmelt flows. The Forks Reservoir primarily redistributes Milk River water within a year, improving Canadian irrigation reliability without increasing total natural flow available for apportionment. Administrative credit mechanisms increase operational flexibility but cannot fully compensate for physical constraints when implemented alone. Combined structural scenarios preserve the individual benefits of storage and conveyance improvements but also reveal persistent system limitations
Integration of Metal-Organic Frameworks with Electrospun Fibers for Heavy Metal Removal and Carbon Dioxide Capture
No full textThe increasing levels of atmospheric CO2 and the widespread contamination of water sources with toxic heavy metals represent two of the most pressing environmental challenges of the 21st century. Efficient and scalable materials capable of capturing CO2 and removing pollutants such as Pb(II) and Cu(II) are urgently needed to mitigate climate change and safeguard public health. In this context, advanced porous materials such as metal–organic frameworks (MOFs) have gained significant attention due to their exceptionally high surface area, tunable porosity, and structural versatility. However, the practical use of MOFs remains limited by their powder form, which complicates processing and application in separation systems. Electrospinning, a versatile and scalable technique for producing nanofibrous membranes with high surface-to-volume ratio, offers a promising strategy to immobilize MOFs in structured and processable forms. This thesis investigates the integration of MOFs with electrospun polymer fibers to develop multifunctional composites for CO2 adsorption and heavy metal removal. Two approaches were explored: direct electrospinning of pre-synthesized MOFs with polymer solutions and in situ growth of MOFs on electrospun mats. For the first time, CALF-20, a Zn-based MOF originally designed for CO2 capture, was incorporated into polyacrylonitrile (PAN) fibers via direct electrospinning. The resulting composites showed high structural integrity with up to 60 wt.% loading and demonstrated excellent removal capacities for Pb(II) and Cu(II), achieving 248.3 mg/g and 128.2 mg/g, respectively. In situ growth of ZIF-67 on electrospun PAN and PVDF fibers was systematically investigated, revealing that solvent type, precursor concentration, and synthesis time strongly influenced particle morphology, crystallinity, and CO2 adsorption performance. Furthermore, the novel in situ growth of CALF-20 on PAN fibers was successfully demonstrated, providing improved accessibility of MOF active sites and enhanced CO2 capture compared to direct electrospinning. Overall, this research introduces new strategies for integrating MOFs with electrospun nanofibers and establishes their advantages over MOFs in powder form or traditional mixed-matrix membranes. By addressing key challenges related to stability, processability, and performance, this work advances the development of scalable MOF–fiber composites as promising materials for next-generation environmental remediation technologies
Countering iron-driven pathology in the CNS with novel neuroprotective analogs
No full textMultiple sclerosis is a chronic inflammatory disease of the central nervous system that leads to progressive neurodegeneration. Among its many pathological features, the accumulation of iron within the brain has gained increasing attention. Despite extensive imaging evidence of iron deposition around and within lesions, especially in deep grey matter, the consequences and the cellular responses to excess iron in the central nervous system remain poorly understood. In this thesis, I investigated how cells of the central nervous system respond to oxidative stress and iron overload. I first characterized the antioxidant enzyme response across several animal models of multiple sclerosis. My results showed that microglia and macrophages consistently upregulated antioxidant enzymes, whereas astrocytes and oligodendrocytes did not exhibit a comparable response. To directly examine the impact of iron in the central nervous system, I developed a new animal model in which ferrous iron, the reactive form of iron, was injected into the white matter of the mouse spinal cord. Unexpectedly, the resulting injury was confined to the adjacent grey matter, which enabled detailed investigation of iron-induced neurodegeneration. I found that neurons exhibited oxidative stress but failed to upregulate protective antioxidant enzymes, while microglia increased several antioxidant enzymes but not glutathione peroxidase 4, the key enzyme responsible for neutralizing lipid peroxidation. These findings suggested that endogenous antioxidant defenses may be insufficient to counter iron toxicity. Therefore, I tested the efficacy of classic ferroptosis inhibitors such as ferrostatin-1 and liproxstatin-1 in protecting neurons against iron. My results showed that, although these drugs were protective in culture, they were not suitable for animal use because of their poor solubility in saline. To overcome this limitation, we designed and synthesized two new analogs with improved solubility, including PZ16044, which demonstrated comparable efficacy to ferrostatin-1 in cell culture. Additionally, treatment with PZ16044 reduced neuronal loss and prevented motor deficits in the iron-injection model. Overall, this thesis provides new insight into the antioxidant responses of the cells in the central nervous system, establishes a model for studying iron-induced neurodegeneration, and introduces a promising therapeutic analog capable of mitigating iron-driven neuronal injury
Dynamic Modeling and Risk Assessment of Carbon Dioxide Sequestration in Saline Aquifers
No full textGeological carbon capture and storage (CCS) in deep saline aquifers is a critical strategy for mitigating climate change and demands robust understanding of subsurface complexities. This thesis advances the knowledge required for secure and effective CO2 geological sequestration by addressing key challenges in fracture characterization, CO2 solubility modeling, and the dynamic influence of fractures on CO2 migration. First, recognizing the critical impact of natural fractures on subsurface fluid flow, Chapter 1 develops and validates a Variable Shape Distribution (VSD) method for modeling a fracture size distribution. Compared to commonly used power law and scale-dependent methods, the VSD method demonstrates a superior fit across wide fracture size ranges, including small and large apertures, which are often truncated or inaccurately represented by conventional approaches. This advancement provides a more reliable estimation of fractal dimensions and fundamental fracture properties, crucial for accurate reservoir modeling. Next, Chapter 2 addresses limitations in CO2 solubility estimation by introducing an improved Henry's law formulation incorporated into a compositional simulator. This refined model accounts for variations in temperature and salinity, crucial factors often simplified in previous studies. Findings highlight that lower salinity significantly enhances CO2 solubility and injectivity, while temperature's impact is more complex, influencing both solubility and CO2 density, which affects buoyancy. This analysis identifies optimal aquifer conditions for CO2 storage with respect to solubility. Finally, Chapter 3 integrates these understandings into a comprehensive investigation of the dynamic effects of various fracture states on CO2 storage performance, employing a coupled geomechanical-compositional simulation model, represented by Barton-Bandis model. Three distinct vertical fracture scenarios were analyzed: confined, permanently open, and pressure-activated. Results demonstrate that confined fractures primarily re-route CO2 migration pathways but have limited impact on an overall plume radius. In contrast, permanently open fractures act as highly efficient conduits, causing significant CO2 leakage. Crucially, pressure-activated fractures, which are more geologically realistic, exhibit a dynamic transition from a barrier to a conduit under elevated reservoir pressure, facilitating leakage that is less severe than from permanently open fractures.
The overarching conclusion is that the physical state of a fracture (confined, open, or pressure-activated) is a more critical factor governing CO2 plume behavior, sweep efficiency, and storage security than its simple presence or proximity to a well. This research underscores the indispensable need for integrating geomechanical coupling in future CO2 storage models to accurately predict dynamic fracture behavior, thereby enabling more robust risk assessments and optimizing long-term sequestration strategies in complex geological environments
Predicting First Lifetime Onset of Internalizing Disorders in High-Risk Youth in the Adolescent Brain and Cognitive Development (ABCD) Study
No full textProspective, early identification of adolescents at imminent risk for their first‐lifetime onset of internalizing disorders is critical for timely preventive care. Leveraging data from youth in the Adolescent Brain and Cognitive Development (ABCD) Study who were free of these disorders at ages 11-12, I developed and compared linear (Elastic Net logistic regression) and non-linear (Random Forest) classifiers using psychosocial assessments and structural MRI metrics to predict the first lifetime onset of DSM-5 depressive, anxiety, or internalizing disorders (either anxiety or depression) by ages 13-14. Analyses were conducted in high-familial-risk, low-risk, and combined cohorts under a nested cross-validation framework. Models trained solely on psychosocial features achieved robust discrimination (AUROC: 0.76–0.81), with negligible improvement from including neuroimaging data (ΔAUROC < 0.01) and similar performance between linear and non-linear algorithms (ΔAUROC ≤ 0.02). SHapley Additive Explanation (SHAP) analyses revealed a consistent set of top predictors across cohorts and disorders, led by sex, behavioral inhibition, sleep disturbance, and parental psychopathology, followed by family functioning, pain, and peer victimization. These findings highlight that a range of psychosocial and health-related factors factors additively predict first-onset internalizing disorders in early adolescence, whereas structural MRI adds little incremental value. Such ML approaches can inform scalable, individualized screening tools to guide early intervention strategies
Perspectives of people who use drugs on implementing overdose response technologies in acute care settings: a qualitative study
No full textAbstract Background People who use drugs (PWUD) face many barriers in healthcare settings. Illicit substance use during hospital stays, including solitary use in bathrooms, is prevalent, leading to a higher risk of overdoses. In this study, we examine the views of PWUD on implementing novel strategies such as overdose response technologies (ORTs) into acute care and explore the perceived acceptability, impacts, and barriers of these interventions. Methods We used convenience sampling to recruit 10 participants from hospitals and addiction medicine clinics, and semi-structured interviews were conducted. The interviews included an explanation of the five main types of ORTs relevant to acute care settings (hotlines, applications, overdose buttons, reverse motion detectors, and wearables), followed by questions where the participant had to critically evaluate whether each ORTs would be effective for each scenario. Open-ended coding and thematic analysis were used, and themes were derived from the data as it was reviewed. Results Participants acknowledged the advantages and potential risks of integrating ORTs into acute care. It was recognized that ORTs could help improve the relationships between PWUD and healthcare providers, reduce mortality rates in hospital bathrooms, and provide peer support during hospital stays. PWUD highlighted privacy concerns, logistical barriers, and stated that ORTs can also negatively impact their relationships with healthcare providers due to stigma. Conclusion Although many participants felt that incorporating ORTs would be an advantage to their care within hospitals, our study also highlighted implementation barriers and broader policy changes that need to be addressed. Working towards addressing such barriers and changes can allow ORTs to be the next tool to help mitigate barriers faced by PWUD
Essays on Decision-Making with Queueing Models for Service Systems
No full textThis thesis consists of three essays that propose simple dynamic time- and/or state-dependent policies to improve the performance of queueing-based service systems. The first essay analyzes a single-physician emergency department (ED) model with heterogeneous new and in-progress (IP) patients over a finite shift with end-of-shift handoff penalties. The system is formulated as a transient optimal control problem, and the analysis shows that the optimal policy has a time-threshold-type structure. For prioritizing between different customer classes, the essay derives time-dependent rules that can be viewed as generalizations of well-known index-based prioritization rules in the literature. It also provides closed-form approximations for the optimal time thresholds as functions of model parameters. Numerical results based on a simulation calibrated with real hospital data show that the proposed policy can significantly outperform current practices and alternative policies in terms of several performance measures, including waiting times, throughput, and handoffs. The second essay extends the first essay to a more general shift-based system with a generic, rather than ED-specific, application and formulates a finite-horizon Markov decision process (MDP). It is shown that the optimal policy is a state- and time-dependent threshold rule on the number of IP customers, and demonstrates that this policy can outperform purely time-dependent rule proposed in the first essay, especially under moderate load. To make the policy easier to use, the essay parameterizes the threshold as a function of time and uses supervised learning to map model parameters to threshold parameters. The third essay studies an infinite-horizon, misclassification-aware MDP for routing patients from emergency medical services (EMS) to either a hospital ED or an alternative urgent care center when only predicted-discharge patients are eligible for diversion and the prehospital classifier is imperfect. It shows that the optimal routing policy has a state-dependent threshold form in the urgent care workload. The classifier design is also parameterized through effort and a propensity threshold. A heuristic method is proposed to solve the joint problem of routing, effort, and classification threshold. Structural analysis as well as numerical experiments show how optimal routing decisions depend on misclassification rates and unit workloads. It is also shown that the optimal effort level is strictly positive but typically small relative to the maximum feasible range
Designing Pores for Proton Conduction in Metal Phosphates, Coordination Polymers and Organic Polymers
No full textHydrogen fuel cells offer a promising route to clean, on-demand energy; however, efficiency and stability challenges have limited their widespread commercial adoption. One of the main issues lies in the sensitivity of the electrolyte of the hydrogen fuel cell which is the material responsible for transporting protons from the anode to the cathode. The current state of the art in electrolyte technology is Nafion®, a family of perfluorosulfonic acid polymers, which are highly conductive but suffer from strong humidity dependence. Since maintaining hydration requires complex water management systems, developing proton exchange membranes with reduced humidity dependence has become a major research focus. This thesis details investigations into several novel proton conductive materials, evaluating their stability, structure and humidity requirements. Chapter One provides an overview of hydrogen fuel cells, the design principles behind proton exchange membrane electrolytes and examples of materials that have achieved success in this field. Chapter Two outlines the specialized methods and characterization techniques employed. Chapter Three investigates the synthesis and conductivity of an amorphous proton conductive material derived from phytic acid, an inositol hexaphosphate ligand. A reproducible synthesis is detailed, a challenge owing to the amorphous nature, along with proton conductivity studies. Upon calcination, the material transforms into a mixture of AlP3O9 and AlPO4, which displays low conductivity. In Chapters Four and Five a 1-D lanthanum coordination polymer was synthesized from a 1-hydroxy-3-(4-imidazolyl)-propane-1,1-bisphosphonate ligand. This coordination polymer demonstrates solvent dependent flexing behaviour and moderate proton conductivity. When soaked in a 5M aqueous imidazole solution, the framework incorporates imidazole molecules in a highly stable, charge compensating manner. This enhances the material’s anhydrous proton conductivity relative to the unloaded form. Chapter Six explores synthetic pathways towards a metal ion templated material. Although the template did not incorporate into the polymer as intended, the reaction yielded a porous polymer upon template removal. Taken together, this thesis provides insights into strategies for designing stable, more humidity independent proton conductive materials for fuel cell applications