Event-Based Histogram Of Gradients For Lane Detection, Ganesh Gupta

Computer Science and Engineering Theses - Archive

In the rapidly evolving landscape of autonomous driving technology, lane detection systems stand as fundamental guardians of vehicular safety. The National Highway Traffic Safety Administration identifies unintentional lane departures as responsible for approximately one-third of all road accidents—a sobering statistic that underscores the critical importance of robust lane detection methodologies. This thesis embarks on an academic exploration at the intersection of neuromorphic engineering and computer vision, examining how the distinctive properties of event-based cameras might be harnessed to enhance lane detection capabilities under challenging environmental conditions. Unlike conventional frame-based imaging sensors that capture entire scenes at fixed intervals, event-based cameras …

Exploring Instruction Generation For Uavs: Dataset Adaptation, Model Behavior, And Diagnostic Insights, Seyedarman Vaziri Bozorg

Computer Science and Engineering Theses - Archive

This thesis explores the development of an answering agent capable of generating natural language instructions for unmanned aerial vehicles (UAVs), grounded in a limited, real-world dialogue dataset. The objective is to adapt a static dataset into a training pipeline that can support instruction generation and serve as a foundation for future interactive systems involving question-asking agents and internal dialogue. A hybrid architecture is implemented using a semantic teacher model (MPNet) and a T5-base encoder-decoder trained with contrastive and supervised objectives. The adapted training process yields statistically acceptable performance across standard evaluation metrics. However, qualitative analysis reveals a mismatch between metric …

Training Data Privacy In Machine Learning: A Systematization Of Attacks And Defenses, Mohammad Sufyaan Saeed

Computer Science and Engineering Theses - Archive

Training and deploying Machine Learning (ML) models introduce significant data confidentiality risks, as modern models can inadvertently memorize and leak information about their training data. While attacks such as membership inference and model inversion are well studied, the literature remains fragmented, with inconsistent threat models and unclear relationships across attack classes and defenses. This work presents a Systematization of Knowledge (SoK) that unifies the landscape of training-data privacy attacks and defenses, aligning them with the NIST Adversarial Machine Learning (AML) taxonomy to enable standardized threat modeling and comparison. Our analysis shows that, despite significant progress in characterizing attack vectors, defenses …

Transformer And Recurrent Architectures For Dynamics Prediction And Policy Learning On Long-Horizon Tasks, Vinal Jitendrabhai Gadhiya

Computer Science and Engineering Theses - Archive

Model-based reinforcement learning promises improved sample efficiency by learning environment dynamics and using them for planning or policy improvement. However, the choice of neural architecture for dynamics prediction significantly impacts the model's ability to capture temporal dependencies and maintain long-term context, capabilities crucial for complex, open-world environments.

This thesis investigates three neural architectures for learning world models: Transformer-based, GRU-based, and a hybrid Transformer+GRU approach. We evaluate these architectures on Crafter, a 2D open-world survival environment that requires long-horizon planning and sequential task completion. In Crafter, agents must perform hierarchical sequences of actions, such as collecting wood, placing a table, and …

Diversity-Driven Xor Secret Sharing: Reliable And Secure Multi-Path Transmission, Richard M. Olu Jordan

Computer Science and Engineering Theses - Archive

The growing reliance on distributed storage and multipath communication sys- tems has intensified the need for security mechanisms that remain robust even when individual nodes or channels are compromised. Secret sharing provides an information- theoretic approach to achieving both confidentiality and availability, and XOR-based constructions in particular offer lightweight and highly structured designs. This thesis develops a unified analytical framework for understanding and evalu- ating XOR-based secret sharing schemes across multiple operational settings, includ- ing plaintext storage, encrypted-data scenarios, and noisy binary symmetric chan- nels (BSCs). Building on a general (t, n) system model, we examine five threshold configurations—(2, 3), …

Crime Theory Informed Agent-Based Modeling For Crime Prediction And Patrolling Route Optimization, Shohreh Moradi

Industrial, Manufacturing, and Systems Engineering Dissertations - Archive

Crime reduction remains a global priority, demanding both accurate modeling of criminal dynamics and efficient allocation of scarce policing resources. To address these needs, this study presents a two‐fold framework that (1) simulates street‐level crime patterns using an agent‐based model (ABM) grounded in Routine Activity Theory (RAT), Rational Choice Theory (RCT), and Crime Pattern Theory (CPT), and (2) optimizes patrol routing through a time-dependent, multi‐visit mixed‐integer linear programming (MILP) formulation.

In the first component, we integrate real‐world crime, environmental, and census data to reproduce realistic offender, citizen, and Police behaviors, capturing where and when robbery, burglary, and larceny occur across …

Guided-Mode Resonant Nanopatterns For Raman Generation And Photonic Devices, Renjie Chen

Electrical Engineering Dissertations - Archive

This dissertation explores advanced strategies for enhancing Raman amplification in silicon photonic devices, focusing on guided-mode resonance engineering and resonant mode manipulation. Silicon, despite its indirect bandgap, exhibits a strong Raman scattering coefficient, enabling it to function as a viable gain medium for integrated photonic systems. However, the realization of efficient, compact, and low-threshold silicon Raman amplifiers and lasers necessitates innovative design approaches that overcome inherent material and structural limitations.

The first chapter provides a fundamental overview of optics, including physical principles, spectral characteristics, guided-mode resonance, simulation methods, and nanopattern fabrication methods.

The second chapter delves into silicon-based Raman amplification …

Data-Driven Forecasting Of Power Demand Via Convex Optimization, Mersedeh Ashraphijuo

Electrical Engineering Dissertations - Archive

This dissertation develops interpretable, data-driven frameworks for short-term power demand forecasting using convex optimization and advanced feature engineering. The models combine historical load data, calendar structures, and meteorological variables to deliver accurate point, quantile, and probabilistic forecasts. By leveraging multi-periodic Fourier features, temperature-based regressors, and autoregressive memory, the proposed approach balances predictive performance with interpretability and scalability. Evaluations on multi-year datasets across US regions show consistent accuracy gains over benchmarks, while preserving transparency critical for real-world deployment. Beyond power systems, the framework generalizes to other time series applications in data science and AI, offering a robust, explainable alternative to black-box …

Characterization And Modeling Of Polyphase Machines And Power Conditioning Components, Nathan Lockhart

Electrical Engineering Dissertations - Archive

Microgrid power configurations have become increasingly prevalent in recent power systems due to the rise of power electronic energy generation, energy storage, and the many diverse electrical demands. Microgrids offer numerous advantages over traditional power electronic networks, which rely on large rotating generators to supply power over extensive distances to multiple users. Remote power grids are particularly beneficial for smaller networks that may be isolated or have unique power requirements, often incorporating energy storage to enhance operational flexibility. Advances in power electronics, such as medium voltage DC distribution, are enhancing the reliability, redundancy, and integration capabilities of isolated microgrids.

To …

Multi-Modal Model-Based Optical Flow Estimation For Event-Based Vision, Pritam Karmokar

Computer Science and Engineering Dissertations - Archive

Event cameras offer a fundamentally different sensing paradigm by asynchronously capturing brightness changes at high temporal resolution, directly encoding motion in the scene. However, their sparse and non-traditional data format poses significant challenges for dense motion estimation, particularly in the context of optical flow. Contrast Maximization (CM) has emerged as a powerful model-based framework for estimating optical flow from event data by optimizing the sharpness of motion-compensated event representations. This dissertation builds upon and significantly advances the CM framework through two complementary contributions.

First, we propose Edge-Informed Contrast Maximization (EINCM), a hybrid approach that augments the traditional events-only CM framework …

An Slo-Aware, Multi-Pronged Approach To Enhancing Resource And Energy Efficiency In Server Applications, Ning Li

Computer Science and Engineering Dissertations - Archive

Server applications operating in oversubscribed cloud environments face the dual challenges of meeting strict Quality-of-Service (QoS) requirements and improving resource and energy efficiency. As the number of user connections and workload diversity continue to grow, existing scheduling mechanisms struggle to balance QoS guarantees, fairness, resource efficiency, and power consumption. This dissertation presents a unified, cross-layer framework to address these challenges through three key contributions: AppleS, UTSLO, and REEF.

First, we propose AppleS, a user-space QoS-aware fine-grained I/O scheduling framework that delivers fair and efficient service to concurrent client connections. AppleS introduces a QoS-centric metric that guides admission control and scheduling …

Synthesis And Application Of Cu-Based Ultrasmall Nanoparticles For Targeted Glioblastoma Treatment, Ryan T. Hart

Material Science and Engineering Dissertations - Archive

Glioblastoma (GBM) is a highly malignant form of brain cancer with a bleak prognosis. Current maximal treatment involves surgical resection followed by chemo-or radiotherapy. Two critical barriers make treating GBM a formidable challenge. First, the tumor's tendril-like proliferation throughout healthy brain tissue often makes complete surgical removal and conventional radiation insufficient. Second, the blood-brain barrier (BBB), which protects the brain, unfortunately also shields GBM from life-saving anti-cancer drugs. Here, we develop a radioactive nanoparticle-based therapy that directly confronts these limitations, offering a new strategy for GBM treatment.

Prostate-specific membrane antigen (PSMA) is a promising target for glioblastoma (GBM) treatment because …

Phase-Field Modeling Of A Protective Layer For The Suppression Of Dendrites In Metallic Anode-Based Rechargeable Batteries, Bharat R. Pant

Material Science and Engineering Dissertations - Archive

Metallic anodes, such as Lithium (Li) and Zinc (Zn), offer promise for the development of high-capacity rechargeable batteries. However, metallic anode-based batteries suffer from severe dendrite growth during the plating process, which poses safety hazards and accelerates capacity degradation. Experimental studies suggest that using a protective coating on the anode surface could potentially mitigate dendrite growth and prolong the battery’s life. However, there are limited theoretical studies on the inhibition effect of a protective layer on dendrite growth. Herein, we developed a phase-field model to simulate the impact of a protective layer on the plating and stripping cycle for Li …

Fabrication Of Microscale Oxide Architectures On Silicon Via A Cmos-Compatible, Deposition-Last Process, Jamal A. Brown

Material Science and Engineering Theses - Archive

This work explores the feasibility of integrating complex oxide devices on silicon using a CMOS-compatible, deposition-last approach. While previous demonstrations of this method have succeeded at larger scales, this study focuses on extending the process to microscale features, with lateral dimensions as small as two microns. The fabrication sequence begins with photolithographic patterning and reactive ion etching of the silicon substrate, followed by the deposition of a silicon nitride mask to delineate device regions. We then epitaxially grew SrTiO3 and La-doped SrTiO3 layers on top of the nitride mask via molecular beam epitaxy. Electrical transport measurements of the La:STO layer …

Improving The Reliability Of Parts Produced Via Laser Powder Bed Fusion Through A Data-Driven Geometry Optimization Methodology, Federico Venturi

Mechanical and Aerospace Engineering Dissertations - Archive

This research aims to qualify the effect of design geometry on quality metrics of additively manufactured (AM) components that define reliability. Through the use of a novel methodology to characterize these quality metrics, AM components are inspected for the presence of defects such as surface roughness notches and porosity. These directly hinder the high-cycle fatigue characteristics demonstrated through the Kitagawa-Takahashi diagram and the El-Haddad model. By demonstrating the effect of geometry through a designed experiment and the adoption of the Murakami square root area parameter and Arola-Ramulu model, the improvement to fatigue life can be quantified. Incorporating this data into …

Stress Analysis Of Anisotropic Inclusion Problems Using Complex Variables, Liming Chen

Mechanical and Aerospace Engineering Dissertations - Archive

This research presents an analytical framework for determining stress fields in an elastic medium containing a circular anisotropic inclusion embedded in an infinitely extended isotropic matrix subjected to far-field uniform stresses. Stress distributions within both the inclusion and the matrix are described using classical stress functions, widely employed in two-dimensional elasticity theory.

To manage the complexity of the mathematical derivations, symbolic computation software (Mathematica) is used to streamline the analysis and obtain closed-form solutions. This approach overcomes traditional computational barriers that have limited the practical application of complex variable methods (CVM) in elasticity.

The methodology builds upon the foundational work …

Design Strategy For Hybrid Thrust Air Bearings: Comparative Analysis Of Rigid Vs. Foil Bearings Considering Optimum Taper Angle And Orifice Location With Experimental Validation, Ehiremen Ebewele

Mechanical and Aerospace Engineering Dissertations - Archive

Gas foil thrust bearings (GFTBs) are contactless bearings that offer advantages such as lightweight construction and the ability to accommodate misalignments and geometric irregularities. However, their load capacity is lower than rigid or magnetic bearings. The structure and geometry of the foil significantly influence GFTB performance. The taper-flat design is the most used configuration due to its effectiveness and ease of implementation. Key parameters in designing this geometry include taper ratio, taper height, orifice size, and orifice location. These parameters must be optimized alongside manufacturing constraints to produce an effective GFTB. This study presents a design optimization investigation of various …

Experimental And Numerical Modelling-Based Optimization Of Additive Manufacturing Processes, Vishnu V. Ganesan

Mechanical and Aerospace Engineering Dissertations - Archive

ABSTRACT

Experimental and Numerical Modeling-Based Optimization of Additive Manufacturing Processes

Vishnu V Ganesan, Ph.D.

The University of Texas at Arlington, 2025

Supervising Professor: Dr. Ankur Jain

Experimental and numerical modeling play a pivotal role in advancing additive manufacturing technologies by enabling a deeper understanding of complex, multi-physics processes that govern part quality, performance, and reliability. These manufacturing techniques—ranging from Powder Bed Fusion (PBF) and Material Extrusion (MEX) to Automated Fiber Placement (AFP)—involve tightly coupled thermal, mechanical, and material phenomena that are challenging to capture through empirical observation alone. Experimental methods offer critical validation and insights into real-world behavior, while numerical …

Intelligent Microfluidic Systems For Precision Manipulation And Real-Time Recognition Via Dielectrophoresis And Deep Learning, Negar Danesh

Mechanical and Aerospace Engineering Dissertations - Archive

This dissertation introduces intelligent microfluidic platforms by combining advanced DEP-based manipulation with real-time visual feedback. A DEP device featuring circular corral traps and dual-plane electrodes enables precise submicron particle trapping, high-resolution particle separation, and cell-particle co-assembly. Simulations and experiments confirm enhanced electric field control and stable confinement. To enable adaptive operation in EWOD systems, a deep learning model (U-Net) was developed for real-time droplet meniscus segmentation. The model achieved 98% accuracy and remained robust under noisy, low-contrast conditions. A live video pipeline was implemented, enabling consistent frame-by-frame feedback for closed-loop control. Together, these innovations establish a foundation for autonomous, high-performance …

Nonlinear Bump Stiffness Model And Its Effect On Structural Stiffness And Nonlinear Rotordynamic Characteristic Of Foil Bearing, Woongeon Lee

Mechanical and Aerospace Engineering Dissertations - Archive

Bump foils are the most widely used in foil bearings, but the behaviors of bump foils are complicated, and their characteristics have been a focus of research for decades. Bump foils are usually modeled as stiffness and damping are accounted for through interactions with shaft eccentricity, loading, shaft speed and excitation frequency. These nonlinear characteristics of the bump foil of radial foil bearings can be observed during both manufacturing and operational processes because of their inherent structural properties such as bump geometry, forming process, age-hardening process and complicated contact behavior with bearing housing. These nonlinear characteristics are one of the …

Multi-Objective Design Optimization Of Hypoid Geared Rotor Systems, Xinqi Wei

Mechanical and Aerospace Engineering Dissertations - Archive

Hypoid gears represent one of the most generalized and complex forms of gearing, widely used for power transmission of skew shafts in vehicles, aviation, and marine transmission applications. Optimizing their performance remains challenging due to the complex tooth surface and contact behavior. Specifically, the design parameters of the tooth surface are multi-scale, interdependent, and subject to strong constraints, leading to strong nonlinearity and an ill-conditioned Jacobian matrix in the parameter identification model. Moreover, feasible and insensitive contact conditions are difficult to constrain due to the inherent complexity of local conjugate contact between the meshing surfaces. These challenges significantly increase optimization …

Computational Study Of Detonation Wave Propagation And Propellant Injection In Detonation Engines, Jayson C. Small

Mechanical and Aerospace Engineering Dissertations - Archive

This research investigates the physics of detonation waves and their application to detonation engines, with two primary objectives: (1) to characterize detonation wave propagation, specifically detonation speed and wavefront thickness, in stoichiometric propane-oxygen mixtures near quenching conditions, and (2) to examine the physiochemical processes during propellant injection under detonation engine conditions, focusing on flush-wall-mounted, fluidic-valve injectors.

A multi-fidelity computational approach was employed. Reduced-order modeling based on the Chapman–Jouguet and Zeldovich–von Neumann–Doring models were used to evaluate detonation parameters and select an appropriate chemical mechanism. High-fidelity, multi-dimensional simulations solved the unsteady, compressible, reacting flows with finite-rate chemistry, using Reynolds-Averaged Navier–Stokes equations …

Topology-Driven Performance Analyses In Consensus Algorithms For Multi-Agent Systems, Brandon Ayala

Mechanical and Aerospace Engineering Theses - Archive

This thesis presents a simulation-based analysis of consensus algorithms in multi-agent systems, focusing on how network topology influences convergence performance. Both first-order and second-order linear consensus dynamics were examined across a variety of network configurations, including undirected and directed versions of cycle graphs, star graphs, minimum spanning trees, and fully connected graphs. In addition to standard consensus problems, leader-follower network structures were introduced to explore the impact of leader placement on convergence behavior, and extensions to multi-leader systems were analyzed to study robustness and convergence under multiple reference inputs. Graph-theoretic properties such as connectivity, degree distribution, and Laplacian eigenvalues were …

Multi-Agent Differential Games Under An Altruistic Equilibrium, Craig Alan Lovell

Mechanical and Aerospace Engineering Theses - Archive

This work studies a multi-agent differential game with linear dynamics under the Berge equilibrium. The governing coupled differential equations for a two-agent and a three-agent game under the Berge equilibrium are derived. These games are simulated and compared to the Nash equilibrium. A sensitivity study is performed which validates that, under some criteria, the Nash equilibrium can be recovered from the Berge equilibrium. Policy fusion between the Berge and Nash equilibrium is explored in a two-agent game. A five-agent game under the Berge equilibrium is simulated and multiple teams of agents in this game are evaluated. Finally, a mixed game, …

Deep Neural Network Models For Heatsink Performance Prediction And Optimization In Single Phase Immersion Cooling: Framework For Future Design Tools And Digital Twin Integration, Braxton J. Smith

Mechanical and Aerospace Engineering Theses - Archive

The rapidly rising computational power of modern computing components combined with the advanced packaging techniques being implemented has resulted in exponentially increasing thermal design powers (TDP) from CPUs and GPUs. Traditional air-cooling methods are approaching their effective cooling limits for many of these components, requiring lower supply air temperatures, higher supply air flowrates, and much larger heatsinks to remain feasible. Transitioning from air-cooling to single-phase immersion cooling offers numerous benefits in thermal performance, data-center size reduction, and energy efficiency. To leverage the merits of immersion cooling, the performance of a given heatsink must be predicted and optimized for best performance …

High Frequency Oscillations As Biomarkers Of The Epileptogenic Zone In Children With Drug Resistant Epilepsy, Lorenzo Fabbri

Bioengineering Dissertations - Archive

Epilepsy surgery stands out as the most effective treatment for patients dealing with focal drug-resistant epilepsy (DRE). Its effectiveness depends on successfully removing or disconnecting the epileptogenic zone (EZ), which is the brain area crucial for seizure generation. The seizure onset zone (SOZ) serves as the best approximation of the EZ; this is the region where most seizures begin, identified through invasive electroencephalographic (iEEG) recordings. However, the unpredictable nature of seizures means they can take hours or even days to occur, consuming valuable human and financial resources. As a result, there is a pressing need for an interictal biomarker that …

Strategies For Enhanced Meg Data Analysis In Clinical Practice And Emerging Frontiers, Pegah Askari

Bioengineering Dissertations - Archive

Epilepsy and dementia are debilitating neurological disorders that pose substantial challenges for patients, caregivers, and healthcare systems. Advances in magnetoencephalography (MEG) and signal processing offer new opportunities to improve diagnostic accuracy, surgical planning, and treatment monitoring. This dissertation presents a unified body of work comprising artifact removal, automated event detection, and deep learning-based biomarker discovery. These approaches collectively enhance the clinical utility of MEG for diverse patient populations.

The first study addresses a significant technical obstacle in the management of drug-resistant epilepsy. Patients receiving responsive neurostimulation (RNS) have historically been excluded from MEG as the data is contaminated by device-related …

Functional Enhancement Of Pancreatic Islets Through Photobiomodulation For Potential Diabetes Therapeutics, Kelli Fowlds

Bioengineering Dissertations - Archive

Islet transplantation is a potential therapeutic route for type 1 diabetic patients facing chronic ketoacidosis and/or hypoglycemia unable to be properly regulated with standard insulin administration. However, successful engraftment is hampered by a multitude of factors. Harvested islets face rapid, substantial degradation due to hypoxia and nutrient depletion once isolated. Transplanted islets are additionally susceptible to the instant blood-mediated inflammatory reaction (IBMIR). This combination of factors leads to more than half of transplanted islets failing to engraft post-surgery. Many areas of research are dedicated to investigating alternative approaches or supplemental treatments to improve the success rate of engraftment and insulin …

Novel Electrophysiological Biomarkers In Pediatric Drug Resistant Epilepsy And Genetic Epilepsy Syndromes, Sakar Rijal

Bioengineering Dissertations - Archive

Pediatric epilepsies, particularly those that are drug-resistant or genetically driven, represent some of the most complex neurological disorders encountered in childhood. Central to their pathophysiology is a disruption in the delicate balance between cortical excitation and inhibition (E/I), often resulting from impaired GABAergic interneuron function. This imbalance manifests as aberrant network dynamics and altered neural oscillations, giving rise to seizures and long-term cognitive impairments. In this thesis, we developed a translational framework to identify electrophysiological biomarkers that (i) assess cortical E/I imbalance and (ii) map epileptogenic zones, with the aim of enhancing diagnosis, guiding surgical planning, and informing therapeutic monitoring …

Vendor-Independent B0 Shimming Framework With Application To Metabolic Mri In Human Gliomas, Mahrshi Jani

Bioengineering Dissertations - Archive

High and ultra-high field MRI and MRSI offer markedly improved signal-to-noise ratio and spectral dispersion, enabling in-vivo characterization of tumor metabolism with unprecedented detail. However, these benefits are tightly coupled to high demands on static magnetic field (B₀) homogeneity, particularly in the brain where susceptibility interfaces near the skull base and paranasal sinuses generate complex, higher-order field perturbations. In glioma patients, additional susceptibility variations arise from surgical cavities, hemorrhage, calcifications, and cystic components, further degrading B₀ homogeneity. As a result, shimming often become the main bottleneck limiting robust, whole-brain spectroscopic imaging and, consequently, our ability to map metabolic …