26400 research outputs found
Sort by
Multi-Domain Machine Learning For Biological Classification: Mallard Classification And Protein Function Prediction
No full textMulti-domain machine learning applications have revolutionized how we understand and predict complex biological phenomena. This dissertation presents novel computational methodologies addressing two critical problems: mallard classification using single-nucleotide polymorphisms (SNPs), and protein function prediction via interpretable topic-aware peptide embeddings. The research focuses on distinguishing mallard populations through SNP data, which are inherently characterized by ultra-high dimensionality. The research uses advanced feature-selection and dimensionality-reduction strategies alongside machine learning classification algorithms to identify minimal, yet highly predictive SNP sets crucial for accurate breed differentiation. This framework demonstrates robust performance with optimal computational efficiency, significantly aiding conservation and breed management efforts. Furthermore, the research project also leverages natural language processing techniques applied to biological sequences, specifically employing enzyme-based sequence fragmentation (e.g., trypsin digestion) followed by embedding with Word2Vec models. Topic modeling (BERTopic) of these peptide embeddings facilitates functional classification (Gene Ontology term prediction), achieving ROC-AUC scores comparable to full-sequence models (98.9% vs. 99%). Notably, topic-derived peptides frequently align with known functional motifs, including ligand-binding sites, underscoring their biological significance and interpretability. Collectively, these studies illustrate the power of machine learning for handling diverse biological datasets, providing accurate predictive models and interpretable insights critical for practical biological discovery and decision-making
Seismic Performance Assessment Of Fatigue-Damaged Cantilever Overhead Sign Structures Using Near-Field Ground Motions
No full textCantilever overhead sign support (COSS) structures are one of the most common features of transportation systems, providing navigational information to motorists. The slender geometry, lightweight design, and welded connection at the base make COSS structure susceptible to fatigue cracking under long term cyclic loading from wind-induced galloping, truck-induced gust pressure and so on. COSS structures are traditionally designed considering nominal wind conditions. However, rising incidents of seismic activity in regions like Dallas-Fort Worth related to industrial injection wells in the Permian Basin have created a serious concern regarding the seismic performance of potentially fatigued-damaged COSS structure. The interaction between already existing fatigue cracks and newly emerging high intensity ground motion from seismic activity remains largely unknown and may cause a serious threat to infrastructure safety. This study addressed the gap, providing a numerical framework considering fatigue damage and seismic fragility. Finite element models were developed in Abaqus for round and multisided pole to simulate the fatigue crack development and stress concentration hot spot. All the loads including dead loads, galloping loads, natural wind gust and truck induced gust loads were applied following AASHTO provisions. To account for the earthquake effect, near field seismic demand was characterized by using ten ground motion records observed with high peak ground accelerations and velocity pulses. Incremental Dynamic Analysis (IDA) was performed for each configuration by gradually increasing the spectral acceleration to understand the progressive nonlinear response behavior of COSS structure up to failure. The findings suggest that base plate thickness of the COSS has major impact in stress concentration factors and seismic fragility. It was found that stress propagation is the result of amplified local stress linked to the thinner base plate substantially reducing the structural capacity. A comparative study between round and multisided pole suggested that multisided pole exhibited higher displacement and fragility threshold with 50% exceedance at Sa ? 1.4g. A similar range of exceedance was observed for the round pole at Sa ? 0.95g. While multisided pole reached displacement limit at higher intensity of ground motion, it exhibited greater uncertainty compared to round pole due to geometric discontinuity and resulting stress irregularity under dynamic loading. In contrast, the round pole displayed smooth stress distribution and predictable IDA response, while reaching displacement threshold at small seismic intensities. Thus, it is imperative to examine the trade-off between deformation capacity and response predictability linked when choosing pole geometry for design of COSS structure. The study provided a comprehensive numerical framework for examining seismic vulnerability of COSS structure examining the influence of geometric properties on stress concentration and stress intensification in addition to failure behavior under seismic loading. The study highlights the need for a holistic approach for evaluation of COSS structure integrating the effects of fatigue damage and near-field seismic demand and provides a pathway forward for informed decision-making and enhancing infrastructure safety
Digital Twin for Real-Time Monitoring and Control of Conveyor Systems Using Flexsim, AI and PLC Integration
No full textModern manufacturing is making significant advancements by innovating and automating most processes. However, a major challenge remains: systems are constantly evolving and becoming more complex to analyze. Fortunately, a powerful tool can help, Digital Twin (DT) technology. This technology enables the analysis and optimization of processes like never before. A Digital Twin is a real-time virtual model of a physical system that continuously up dates with live data. One of its greatest features is the ability to create infinite scenarios, allowing hundreds of configurations to be tested virtually, risk-free, and without making any real-world changes that could disrupt ongoing operations. One of the best software solutions for implementing Digital Twins in manufacturing is FlexSim, which specializes in modeling manufacturing and healthcare systems. FlexSim enables users to recreate, analyze, and predict large-scale models. With its real-time analysis tools, users can track performance, utilization, and overall system efficiency with precision. This project developed a Digital Twin (DT) of a conveyor system using FlexSim simulation software, the FlexSim Emulation module, a vision system, and a programmable logic controller (PLC). The system integrates multiple sensors, including a vision sensor, to capture real-time data. The vision system utilizes artificial intelligence (AI) to train a model for object classification. By leveraging a convolutional neural network (CNN) called EfficientNetB0, we created an object classifier capable of identifying object colors and shapes, as well as detecting defective items that the system is not designed to process. A mini-scale prototype was built as a testbed, demonstrating real-time monitoring, re mote oversight, and improved system control. The DT framework is highly scalable, offering significant potential for expanding automation and intelligence in manufacturing. The scholarly contribution of this work lies in the development of a low-cost, scalable Digital Twin framework that integrates AI-driven visual inspection, PLC control, and real-time emulation using FlexSim. Unlike traditional inspection systems, this solution enables near-instant object classification, autonomous decision-making, and full synchronization with physical systems, demonstrating measurable improvements in throughput, inspection speed, and defect handling efficiency
Ghrelin and Decision-Making: Exploring its Role in Food Reward, Novelty-Seeking, and Cost Valuation
No full textWhen low energy levels are detected, Ghrelin, a peptide hormone, is produced in the fundus of the stomach. Referred to as the hunger hormone , Ghrelin is traditionally associated with increased food intake and blood sugar levels. A key modulator in the gut-brain interaction, Ghrelin\u27s growth hormone secretagogue receptor can be found widely across the cortico-striosomal circuit. Critical for decision-making, Ghrelin-based manipulation of this circuit can directly affect reward and cost valuation. While Ghrelin is uniformly considered to increase the valuation of food-based rewards, little research has explored its impact on non-food-based rewards or its effects on cost valuation. Therefore, to examine Ghrelin\u27s effects on behavioral decision-making, rodents performed a series of fifteen behavioral tasks categorized into three contexts: food reward, novelty-seeking, and cost. Notably, in food-based tasks Ghrelin did not increase food-seeking but altered preference patterns, suggesting its role in modulating reward salience over hunger- driven motivation. Ghrelin did, however, increase novelty-seeking behavior in novelty tasks. This preference for novel rewards was retained even in the presence of well-known and novel food rewards, suggesting Ghrelin plays an active role in regulating exploratory behavior. Indicating its role in cost valuation, Ghrelin consistently increased cost aversion and competitive behaviors. A culmination of these results questions the traditional view of Ghrelin as a hunger signal. As demonstrated here, while Ghrelin drives reward-seeking, its effects are highly context-dependent and potentially alter risk sensitivity and persistence in goal-directed behavior. Elucidating Ghrelin\u27s effects on decision-making has vast implications in studying the gut-brain interaction and its impact on metabolic or psychiatric disorders. Disorders such as depression, stress, obesity, and substance abuse commonly use overlapping circuitry and demonstrate altered reward processing, shifted cost valuation, and modified expression of Ghrelin. Future research aims to explore the neuronal mechanisms by which Ghrelin perturbed decision-making to better understand potential therapeutic targets for these disorders
Crossing Borders: Analyzing the Factors Contributing to Healthcare and Reproductive Preferences Among Residents of the Texas-Mexico Border
No full textCrossing Borders: Analyzing the Factors Contributing to Healthcare and Reproductive Preferences Among Residents of the Texas-Mexico Border explores the complexities of healthcare access and utilization among border residents, particularly in the context of reproductive health. Through qualitative analysis of focus group discussions with promotoras - community health workers - this study identifies key factors influencing cross-border healthcare decisions, including economic barriers, cultural influences, and perceptions of quality. The findings reveal that high costs and limited insurance coverage in the U.S. drive many individuals to seek affordable healthcare and medications in Mexico, where pharmacies provide immediate, low-cost services without the need for prescriptions. Additionally, the study highlights the role of community networks in disseminating health resources and information, as well as the challenges faced by undocumented individuals in accessing care. By centering the voices of promotoras, this research contributes to a deeper understanding of the healthcare needs of border communities and offers policy implications aimed at improving access to reproductive health services and fostering binational healthcare collaborations. Ultimately, the thesis underscores the necessity of addressing systemic inequalities in healthcare access along the Texas-Mexico border to better serve these marginalized populations
Retoriando La Neta : Cosechando Testimonios del Campo y Desahogos Laborales as Regenerative Rhetoric Pathways for Raza Obrer@ Performing La Talacha in Marginalized Laboral Spaces
No full textEmic, phenomenological, and interdisciplinary, this study draws on and seeks to bring the fields of Sociopsychology and Rhetoric and Writing Studies into conversation to address the potential intracultural cognitive bias of raza obrer@ (RO), by raza non-obrer@ (RNO), throughout the Borderland (the El Paso/Juarez Region) through the strategic application of Regenerative Rhetoric.
In this dissertation, I examine testimonios del campo (personal narratives from laboral spaces) and desahogos laborales (emotional laboral releases) of RO performing la talacha (arduous blue-collar work) within marginalized labor spaces utilizing Regenerative Rhetoric reparative discourse/dialectic strategies that spur sentipensamiento en la sudoracion (sensing and thinking within the emotional, laboral environment) to cosechar (harvest), retoriar (rhetorically navigate to synthesize) and, ultimately, arrive at la neta (the unbridled, organic truth of the oppressed).
The findings of this study emphasize an existing intracultural cognitive bias amongst gente raza of/about RO that can be initiated to heal with Regenerative Rhetoric as a multidisciplinary application that allows rhetorical scholars a means to traverse into other academic fields to unmute and lift oppressed obrer@ voices. This study also highlights how rhetorical scholars can utilize Regenerative Rhetoric, by retoriando la neta, to better understanding the denigrating effects that social stratification (as a hierarchal framework of power and privilege) and cognitive dissonance (an internal sociopsychological discord) have on on the sociocultural values of gente raza share within/because of la obra (blue/brown-collar trades). This study also serves as a springboard for other viable veins of research that further Working-class Rhetoric investigations within the fields of Rhetoric and Writing Studies
Inference Per Joule: A Performance Metric For Artificial Intelligence In Space Applications
No full textThe use of artificial intelligence (AI) has grown exponentially in recent years. This growth is driven in part by the significant advancements in computing capabilities, which have also increased exponentially. Computers have not only become more powerful but also smaller in size, thanks to the evolution of transistor technology. These developments have enabled AI to become a widely accessible tool, even in recreational activities such as image creation and entertainment videos.
More recently, the use of AI has extended to space applications, where it can enhance and optimize various tasks. However, space conditions pose significant challenges for conventional computers due to the lack of atmosphere, exposure to high levels of radiation, shock and thermal stresses.
This raises an important question: how can we measure the efficiency of a computer running AI in space? For an image processing AI model, a relevant metric is the number of inferences it can process per second. However, in the context of space, limited power consumption becomes critical, as the amount of heat generated - and the subsequent need for dissipation - directly impacts a computer\u27s operation.
To address these challenges, this work proposes a metric that combines these two factors to determine the number of inferences a compute module can process per joule of energy consumed. This metric, referred to as efficiency, serves as a key indicator of performance in space environments. To validate this concept, the following work summarizes three years of research dedicated to demonstrating the feasibility and utility of this metric
Evaluating Extrusion Model Accuracy Using Laser Profilometry In Ceramic Robocasting Via Custom G-Code
No full textThis thesis presents an experimental validation of a simplified extrusion model intended to predict dimensional accuracy in ceramic Direct-Ink Writing (DIW). Given the inherent complexity of ceramic paste formulations, extrusion processes, and shape retention during additive manufacturing, it is crucial to validate simplified models that can accurately forecast bead geometry and subsequent dimensional fidelity. A custom-designed force tester was developed to emulate ram extrusion conditions and determine key parameters using the Benbow-Bridgwater equation while attempting to be a low-cost alternative. This approach is meant for rapid, affordable assessment of ceramic paste extrusion characteristics without the need for complex rheological testing. The force tester fell short of expectations due to the number of tests necessary to acquire the paste characteristics of a single formulation. Single-line, single-layer, and multi-layer prints were produced and assessed using a laser profilometry scanning system. Bead and layer geometries were measured, evaluating the predictive accuracy of the simplified volumetric extrusion calculation typically employed in polymer extrusion slicers. Experimental results demonstrated that the simplified model accurately predicted bead width and height to within approximately 5-10% of targeted dimensions under stable extrusion conditions. Through these investigations, the study identified critical processing parameters that influence dimensional fidelity, including optimal nozzle dimensions, infill overlap and layer spacing. Additionally, the presence of extrusion defects such as bead discontinuities, pressure fluctuations, and liquid-phase migration were characterized and related directly to paste formulation and printing parameters.
Ultimately, this research confirms the practical utility of simplified extrusion models in ceramic DIW, highlighting their capabilities and limitations. The insights gained provide a solid foundation for future advancements in slicer development and extrusion control, aimed at enhancing the predictability, consistency, and dimensional accuracy of printed ceramic parts
Establishing And Advancing Qualification Strategies For Aerospace Components Produced Using LPBF
No full textThe adoption of Laser Powder Bed Fusion (LPBF) for aerospace hardware requires qualification strategies that ensure repeatable, reliable, and certifiable performance. This dissertation focuses on a qualification effort for an aerospace titanium bracket to be used for a serial production aircraft, following a fixed-process, sample-centric methodology. The work was performed following industry expectations for mechanical performance, dimensional tolerance, and process control, and culminated in the successful qualification of the bracket for flight hardware installation.
The qualification effort was divided into four primary objectives. Machine and parameter qualification was first to establish a baseline LPBF process on an SLM 280 HL system using Ti-6Al-4V powder feedstock. Results across repeated builds revealed spatial variations in mechanical performance, particularly in areas with insufficient gas flow. These findings led to a redesign of the gas flow system in collaboration with the OEM, which significantly improved build uniformity. Second, material, process, and supplier qualification were conducted using an extensive test matrix, which included tensile, shear, fatigue, fracture toughness, metallography, and chemistry assessments, all done at NADCAP-certified laboratories. All results met or exceeded aerospace OEM requirements, and the process was frozen for production use.
Third, part qualification confirmed that the LPBF bracket satisfied all functional, dimensional, and material criteria. The first article underwent CMM-based dimensional inspection, tensile testing, and X-ray computed tomography, all of which passed the necessary quality metrics. Additional blue-light scanning during development revealed as-built distortion, informing optimization of build orientation and support strategy. Lessons learned included the impact of hot isostatic pressing on final geometry and the need to lock orientation early to avoid requalification. Fourth, recurring production testing monitored quality over time. Statistical process control tools (Cp, Cpk, Pp, Ppk) revealed that oxygen contamination (due to inconsistent purge practices) led to chemistry drift and degradation in ductility. A change was implemented to ensure purging reached a verified oxygen threshold before each build. Additionally, a comparative fatigue study demonstrated the feasibility of ultrasonic fatigue testing as a rapid, cost-effective screening tool. Equivalent results were achieved between traditional uniaxial fatigue (20 Hz) and ultrasonic testing (20 kHz), with a reduction in test time from almost six days to under ten minutes for 10 million cycles. This finding supports the integration of high-throughput fatigue methods into future qualification protocols. The dissertation concludes with a forward-looking proposal to evolve from a sample-centric to a subsystem-centric qualification model. Rather than relying solely on sample testing, this approach emphasizes the performance of core machine subsystems (mainly gas flow and atmosphere control). Additional recommendations include subsystem benchmarking, beam profiling, powder reuse management, and advanced fatigue characterization.
In summary, this work not only qualifies a functional aerospace component but also contributes a robust case study in the maturation of LPBF process control and lays the groundwork for scalable, repeatable qualification methodologies that will support the broader industrialization of metal AM in aerospace