Engineering Commons^™

Optimizing Integrated Access And Backhaul Topology Using Monte Carlo Tree Search, Srimanth Reddy Ummenthala

Master's Projects

Integrated Access and Backhaul (IAB) plays a central role in enabling scalability and high-throughput in wireless networks, especially in areas where wired backhaul is not feasible. Optimizing the topology of IAB networks is a complex task, involving trade-offs between path loss, node capacity, and link quality. To address this, this project investigates a Monte Carlo Tree Search (MCTS)-based approach to improve overall network performance by maximizing the capacity of network, considering Signal-to-Noise Ratio (SNR) on wireless links. MCTS provides a guided search mechanism to construct topologies efficiently under the constraints and is evaluated against baseline topologies. Experimental results show that …

Dynamic Pricing For Revenue Maximization In 5g Networks With Elastic Network Slicing And Reinforcement Learning, Naisheel Shah

Master's Projects

Network slicing is a concept that enables a diverse array of network applications with lots of unique service requirements. For the scope of this research, we delve into elastic network slicing, which has a potential benefit for both the providers and users through cost-effective resource utilization. Dynamic pricing of these slice resources is a method for operators to realize the balance of different types of network slices by implicitly communicating the current network state to slice users. We have designed a custom pricing scheme using Deep Reinforcement Learning for elastic network slices that maximizes the revenue of slice providers while …

Link Failure Localization In Hierarchical, Multidomain Optical Networks, Martin Mihailov Bojinov

Master's Projects

Optical networks transport data encoded on light signals over optical fiber cables. Individual optical networks are managed by domain administrators, such as service providers, vendors, or regional bodies. Multidomain optical networking explores the possibility of enabling seamless data transmission across domain boundaries. In this project, we explore how we can perform network monitoring in a hierarchical multidomain optical network while preserving domain security and autonomy. This is done by allocating dedicated monitoring trails across various broker abstractions of our network topologies. We propose three heuristic functions that expedite trail selection. Of the three heuristics, our least monitored algorithm performs the …

Performance Comparison Of Machine Learning Across Metal, Cuda, And Neuromorphic Frameworks, Ryan Saini

Master's Projects

Machine learning’s computational demands necessitate optimal performance and utilization. This research compares Apple Silicon M3 Pro with MPS, NVIDIA RTX 3070 GPU with CUDA, and neuromorphic computing for machine learning methods. We provide a cross-platform and cross-architecture performance analysis of machine learning methods to identify optimal configurations for training and inference scenarios. On traditional neural networks, Apple Silicon with MPS delivers superior energy efficiency at the cost of longer processing times for training and inference. NVIDIA with CUDA offers faster computation in training and inference at higher energy costs. Convolutional spiking neural networks perform competitively on event-based data, particularly on …

Suicidal Ideation Detection On Reddit Using Llm-Annotated Data And Graph Neural Networks, Ikbal Singh Gurdev Singh Dhanjal

Master's Projects

Suicide is the fourth leading cause of death among people aged 15-29. More than 720, 000 people commit suicide every year. During the COVID-19 pandemic, we saw an increase in people seeking out mental health support on anonymous forums like Reddit. These anonymous forums allow people to express their suicidal ideation without judgment and give them a support structure that not everyone has. The aim of this project is to detect suicidal ideation using Reddit. In this work, we propose SIRGEL (Suicidal Ideation on Reddit using Graph Embeddings and LLMs), a dual-pipeline approach that combines large language model (LLM)- based …

Malware Opcode Embedding And Quality Assessment Of Generative Sample Embeddings, Atishay Jain

Master's Projects

Malware is software used to damage and disrupt computer systems with the intent to cause damage to the victim. Malware detection and classification into malware families is a crucial problem for cybersecurity researchers. One of the major bottlenecks in improving these systems is the shortage of good quality labeled malware data, especially for malware families with scarce samples. Researchers have utilized generative models to generate malware data to address this issue. Malware embeddings encode patterns within a malware file, which can be used to detect and classify malware. Recently, encouraging results have been obtained in generating malware embeddings using generative …

Transforming Chatbot Text: A Sequence-To-Sequence Approach To Human-Like Text Generation, Natesh Reddy

Master's Projects

With the advancements in Large Language Models (LLMs) such as ChatGPT, the boundary between text written by a human and that created by AI has become blurred. This poses a threat to systems designed to detect AI-generated content. In this work, we adopt a novel strategy to adversarially transform GPT-generated text using sequence-to-sequence (Seq2Seq) models, to make the text more human-like. We concentrate on improving GPT-generated sentences by including significant linguistic, structural, and semantic components that are typical of human-authored text. The goal is to then use this transformed data to train a more robust detection model. Our experiments show …

Human Vs. Llm: Interpreting Emotion From Neutral Faces In Contextually Charged Backgrounds, Sai Mounika Peteti

Master's Projects

In this study, we investigate the intersection of cultural context, visual perception, and implicit bias in interpreting neutral human expressions. Specifically, we explore how emotionally charged backgrounds can shape viewers interpretations of neutral facial expressions. Using an experimental setup, participants are shown neutral human portraits paired with varying background types ranging from emotionally neutral or pleasant to evocative scenes. Participants are tasked with selecting which background best matches the emotional state of the person depicted in the portraits. Eye-tracking data is collected to analyze visual attention patterns and cognitive processing. In parallel, we also evaluate how large language models (LLMs) …

Clustering Organ Cell Types, Venkata Satya Swathi Mattaparthi

Master's Projects

The Human Cell Atlas (HCA) created a reference map of all human cells. My project uses the Tabula Sapiens dataset, developed under HCA and based on single-cell RNA sequencing data, to explore cell type and tissue diversity. I performed experiments using the Elbow method and a formula based on dataset observations to determine the number of clusters, then applied k-means clustering on two representative subsets of the All Cells dataset. Clusters were selected for analysis using Shannon’s Diversity Index and Pielou’s Evenness. A novel algorithm based on the cell differentiation tree was used to validate the biological coherence of the …

Comparative Analysis Of Adversarial Permeability In Cpu-Native, Qat And Onnx-Based Quantized Transformer Models, Fahad Siddiqui

Master's Projects

This work explores securing and optimization of Transformer-based time series forecasting models. We employ several quantization techniques, including quantization-aware training (QAT), and tested the robustness of quantized models by adversarially attacking them. The preliminary results of this work, in our controlled setup, indicate that quantized models outperform the full precision model in terms of robustness against adversarial attacks. They achieved this robustness while showing a very minimal decrease in their forecasting performance.

Framework For Identity Privacy Through Gender Based Skeletonization, Harrison Hwang

Master's Projects

The protection of one’s privacy and sensitive information is becoming increasingly difficult in the modern age full of surveillance and data collection. Through the use of image based object detection machine learning models trained for human and facial recognition, people can be identified and tracked to a terrifyingly accurate degree. On the other hand, the information present in surveillance media can play a key role in security and law enforcement. This presents a problem of how to preserve key information without compromising the privacy of any individuals present in the video. In this research project, Computer Vision techniques and a …

Enhancing Code Review Automation With Large Language Models Using Qlora Fine-Tuning And Rags, Sumukh Naveen Aradhya

Master's Projects

In this technological era where Artificial Intelligence and Machine Learning are revolutionizing various domains, Large Language Models (LLMs) are emerging as a very powerful tool. In the software development lifecycle, it is extremely important to have reliable code reviews to ensure security and maintain code quality. This project aims to survey various existing methodologies to aid creation of efficient code review automation agents and also research on ways to make this process more efficient. Parameter Efficient Fine-Tuning (PEFT) methodologies such as LoRA and QLoRA have been explored with an additional focus on a hybrid model that combines adaptive QLoRA with …

Large Language Models For Bacterial Genomic Analysis, Manvendra Chavan

Master's Projects

Identification of bacterial gene sequences with agricultural applications has the potential to transform agricultural biotechnology. These genes can be used in environmentally friendly pest control strategies. One such use case is identifying genes with potential insecticidal properties. With an increasing number of genomic information and decreasing numbers of available annotated sequences, finding new insecticidal genes has become more challenging.The traditional methods relying on sequence alignment and annotated databases are not effective in detecting functionally relevant genes lacking close homology to known cases. This project investigates the data-driven classification of genes by sequence modeling. This research is focused on learning DNA …

Genegate: Genetic Gating In A Mixture-Of-Experts For Real-Time Multi-Objective Traffic Signal Control, Rashmi Vishwanath Bhat

Master's Projects

Urban traffic signal control often needs to juggle between competing goals. It needs to minimize delays, reduce emissions, prevent crashes, and prioritize emergency vehicles all while the demand is constantly fluctuating. Traditional fixed-time or statically blended policies cannot reallocate priorities quickly when conditions change. We introduce GeneGate, a mixture-of-experts framework that uses a lightweight genetic gate to fuse four specialist controllers (throughput, emissions, safety, emergency) and adjusts their weights in real time. A short offline genetic search produces a robust initial blend, and an online micro-evolution step refines it every few cycles based on live traffic feedback. GeneGate’s adaptive gating …

Advanced Knowledge Extraction With Biomedical Data Using Llms, Akshat Krishna

Master's Projects

The rapid growth of biomedical research has led to an overwhelming volume of unstructured textual data in the scientific literature. This has necessitated the development of an automated approach for knowledge extraction and integration. In

this project, we present a comprehensive pipeline for constructing a unified biomed- ical knowledge graph by combining two well-known datasets: CHEMPROT [1],

which captures chemical–protein interactions, and EU-ADR [2], which annotates drug–gene–disease relationships. In order to identify important biomedical entities and interactions from CHEMPROT dataset, we perform Named Entity Recognition (NER) and relation Extraction (RE) using state-of-the-art biomedical models like BioBERT [3], BioGPT [4] and …

Hydrogen Permeation In Iron-Chromium-Aluminum (Fecral) Alloys And The Effects Of Microstructure And Surface Oxide, Nathan T. Gehmlich, Thomas F. Fuerst, Hanns Gietl, Chase N. Taylor, Joshua Rittenhouse, Haiming Wen, M. Nedim Cinbiz

Materials Science and Engineering Faculty Research & Creative Works

Iron–chromium–aluminum (FeCrAl) class alloys are candidates for use as cladding for accident-tolerant fuels and moderators. In this context, hydrogen isotope permeation in FeCrAl alloys is an important material property. In the present work, the apparent permeability, effective diffusivity, and apparent solubility of hydrogen in the FeCrAl alloys C26M and Kanthal D (KD) were measured with gas-driven hydrogen permeation. Permeation measurements were conducted at temperatures of 400 to 700 °C and at gas-driven pressures from 1 to 100 kPa. In particular, the effect of grain size on hydrogen transport was studied with KD samples with three different microstructures: nanocrystalline (NC), ultra-fine …

Dual-Bed Radioiodine Capture From Complex Gas Streams With Zeolites: Regeneration And Reuse Of Primary Sorbent Beds For Sustainable Waste Management, Krista Carlson, Brian J. Riley, Joshua Turner, David C. Cantu, Dev Chidambaram, Charmayne Lonergan, Jeffrey D. Rimer

Materials Science and Engineering Faculty Research & Creative Works

No abstract provided.

Degradation Of Radioactive Waste Encapsulation Of Bao−Feox−P2o5 Glass Evaluated From Electrical Conductivity, Kazuki Mitsui, Nobuyasu Nishioka, Hiromichi Takebe, Richard K. Brow, Akira Saitoh

Materials Science and Engineering Faculty Research & Creative Works

We report on the electrical conductivity of barium iron phosphate glass, a candidate for radioactive waste encapsulating glass. The glass composition has been optimized for slow dissolution in hot water because the precipitation of a stable hydration layer of FeO(OH) on the glass surface enhances the water durability of the glass. The mechanism of electrical conductivity of the glass is assumed to be an electron hopping between the electronic energy levels of Fe (II) and Fe (III) ions, obeying the electron carrier generation of Fe (II) → Fe (III) + e−. The electrical conductivity of the base glass was ∼3.2 …

Designing Robust Quasi-2d Perovskites Thin Films For Stable Light-Emitting Applications, Sharmistha Khan, Reshna Shrestha, Mengru Jin, Doyun Kim, Guan Lin Chen, Ruipeng Li, Yijia Gu, Qing Tu, Namyoung Ahn, Wanyi Nie

Materials Science and Engineering Faculty Research & Creative Works

Quasi-2D perovskite made with organic spacers co-crystallized with inorganic cesium lead bromide inorganics is demonstrated for near unity photoluminescence quantum yield at room temperature. However, light emitting diodes made with quasi-2D perovskites rapidly degrade which remains a major bottleneck in this field. In this work, It is shown that the bright emission originates from finely tuned multi-component 2D nano-crystalline phases that are thermodynamically unstable. The bright emission is extremely sensitive to external stimuli and the emission quickly dims away upon heating. After a detailed analysis of their optical and morphological properties, the degradation is attributed to 2D phase redistribution associated …

Ground Testing Of A Magnetic-Electrostatic Separation System For Lunar Regolith Beneficiation, Peter Bachle, Charles Wood, Jeffrey Smith, Fateme Rezaei, David Bayless, William Schonberg, Daoru Han

Materials Science and Engineering Faculty Research & Creative Works

The separation of lunar regolith by mineral composition and size category is useful forin-situ resource utilization (ISRU). The research presented herein discusses the designing and development of equipment that has the potential to separate lunar regolith into aluminum, iron-titanium, and magnesium-iron ores. Along with the metal ore separation, this equipment shows the potential to separate regolith by size categories. The combined effect of these separation methods generates output that is valuable to subsequent use in metal and oxygen extraction, additive manufacturing, and regolith sintering processes. The designed equipment uses a dual-strength magnet system with N42 and N52 neodymium magnets for …

The Effect Of Carbon Equivalent And Nodularity On Multi-Axial Casting Wall Movement During Spheroidal Graphite Iron Solidification And Cooling, Noah J. Brack, Mingzhi Xu, Jingjing Qing, Simon Lekakh

Materials Science and Engineering Faculty Research & Creative Works

Spheroidal graphite iron, also known as ductile iron, is an iron–carbon casting alloy used in industry for its good castability, balanced mechanical properties, and low cost. Ductile iron consists of round graphite nodules in an iron matrix. During solidification and cooling, ductile iron castings experience dynamic volume changes due to the precipitation of graphite nodules and formation of austenite. These dynamic volume changes can distort external casting surfaces, causing swell and shrinkage porosity. A novel apparatus was custom-built to capture the casting wall movement in real time along three axes. This study found that casting expansion increased with carbon equivalent …

A Novel Technique For Improved Measurement Of Graphite And Inclusions In Ductile Iron Contaminated By Boron, Chase Schroeder, Colleen Lehrer, Simon Lekakh, Laura Bartlett

Materials Science and Engineering Faculty Research & Creative Works

The highly heterogeneous microstructure of ductile iron in casting includes graphite particles, non-metallic inclusions, microporosity, and other features, such as carbides distributed in the ferrite–pearlite matrix. Determination and comprehensive quantification of microstructural features are practically important however challenging. An advanced methodology based on an automated SEM/EDX analysis was developed and compared to the standard optical imaging method. In addition to backscattered electron contrast, sensitive to atomic number, a novel methodology added EDX data to identify and classify the multiple structural features at micron resolution thresholding. To quantify the shape of individual phases, the eight chord raster algorithms was used and …

Evaluation Of Processing Conditions For The Reduction Of Electrochemical Salt Waste Using Phosphate-Based Dechlorination, Harmony Werth, Jade Beland, Paige Murray, Dave Liang, Laurel Sharpless, Jonathan Evarts, Charmayne Lonergan, Brian J. Riley, Michael Simpson, Krista Carlson

Materials Science and Engineering Faculty Research & Creative Works

Electrochemical processing of used nuclear fuel in molten chloride salts generates complex radioactive salt waste. Dechlorination of waste salt using phosphate compounds at elevated temperatures (∼600 °C) reduces the volume of material for disposal while evolving gaseous chlorine compounds that could be used to transform metallic uranium into UCl3. In this study, the effects of processing temperature, environment (air or argon), and H3PO4 precursor-to-chlorine ratio on the dechlorination efficacy of a simple alkali salt mixture (SSM) and a salt waste simulant (ERV3) were evaluated. For both salts, the highest chlorine release was concurrent with the water boiling. For the SSM, …

Controlling Nitrogen Pickup During Induction Melting Of Ultrahigh-Strength Cr-Ni-Mo-V Steels, Kingsley T. Amatanweze, Viraj A. Athavale, Mario F. Buchely, Laura N. Bartlett, Ronald J. O'Malley, Daniel M. Field

Materials Science and Engineering Faculty Research & Creative Works

Nitrogen pickup during air induction melting can result in porosity and a loss of fracture toughness in ultrahigh-strength quenched and tempered steel castings. Nitrogen atoms are easily adsorbed into liquid steel upon exposure to the air, and argon shrouding alone has limited effectiveness. Previous studies have shown that proper charge sequencing and maintaining a high amount of dissolved oxygen in the melt prior to tapping and deoxidation can limit nitrogen pickup in the melt. In the current study, the effect of melt practice and charging procedure on nitrogen pickup was studied as a function of hold time in a series …

Zrc Particle Size Effect On The Mechanical Properties Of Spark Plasma Sintered 60–40 Vol% Zrc–Mo Cermets, Nathaniel Blatt, Jeremy Watts, Brian Taylor, Jhonathan Rosales, Gregory Hilmas

Materials Science and Engineering Faculty Research & Creative Works

Cermet's of 60 vol% ZrC and 40 vol% Mo with different starting ZrC particle sizes were spark plasma sintered to full density at 1950 °C. Elastic moduli, Vickers hardness and fracture toughness were measured at room temperature and flexural strength was measured up to 1600 °C. Young's modulus was 365 ± 5 GPa, shear modulus was 149 ± 2 GPa, Poisson's ratio was 0.22 ± 0.01, and no significant difference was observed for the moduli with respect to the starting ZrC particle size. Room temperature flexural strength was affected by the ZrC particle size, measured as ∼524 and ∼598 MPa …

Parametric Analysis Of Water Jet Descaling Efficiency Of Reheated Continuously Cast Thin Slab, Tochukwu P. Ojiako, Mario F. Buchely, Simon Lekakh, Ronald J. O'Malley, Richard Osei, Taha Tayebali

Materials Science and Engineering Faculty Research & Creative Works

Efficient oxide scale removal is critical for maintaining surface quality and process efficiency in steel manufacturing. This study optimizes water jet descaling by evaluating the performance of flat and rotary jet nozzles under varying process parameters. Using a combined approach of experimental analysis and computational fluid dynamics, it investigates the influence of pressure (138–275 bar), lead angle (0°, 15°, 25°), working distance (50–100 mm), and spray angle (15°–25°) on descaling efficiency. Findings indicate that flat jet nozzles achieve superior performance at short working distances due to concentrated impact forces, while rotary jet nozzles sustain efficiency over extended distances through dynamic …

Ultrafast Belite (Β-C2s) Production Using Biofuel-Based Combustion Synthesis, Shubham Agrawal, Sayee Srikarah Volaity, Oluwadamilare Charles Adesina, Bryan K. Aylas-Paredes, Taihao Han, Srinivas Kilambi, Aditya Kumar, Narayanan Neithalath

Materials Science and Engineering Faculty Research & Creative Works

This study presents a novel low-temperature, combustion synthesis (CS) approach for the rapid production of belite-rich cements. CS leverages the exothermic heat released from the combustion of biofuels such as lignin and/or biomass, intermixed with pelletized limestone and quartz. At an imposed furnace temperature of ∼700°C (as opposed to 1200°C–1300°C required in a conventional kiln), the source materials are rapidly transformed to belite, resulting in energy, emissions, and economic benefits. This work explores the influence of various process parameters, viz., fuel types and contents, airflow rate, porosity, holding temperature, and holding time on the efficiency of CS-based belite synthesis. Through …

Roadmap To 100 Gwdc: Scientific And Supply Chain Challenges For Cdte Photovoltaics, B. Edward Sartor, Samantha B. Reese, Lorelle M. Mansfield, Ryan Muzzio, Chun Sheng Jiang, Eric Colegrove, John S. Mangum, Camden L. Kasik, Daniel Z. Shaw, James R. Sites, Ramesh G. Dhere, Alex B. Goldstone, Michael S. Moats

Materials Science and Engineering Faculty Research & Creative Works

This roadmap highlights pathways to expand the CdTe module manufacturing capacity per year to 100 GWDC by 2030 by improving Te extraction from existing supply chains, minimizing Te usage in modules by leveraging thinner absorbers, and focusing research efforts in key areas to improve module efficiencies. Both scientific and supply chain innovations will be necessary to maintain the high compound annual growth rate of the CdTe photovoltaic (PV) industry and cement its role as a key technology for multi-TW-scale PV deployment.

Leaning Pine Arboretum Gazebo, Carson C. Henry

Construction Management

The Leaning Pine Arboretum gazebo provides an aesthetic relaxation space for visitors to the arboretum as well as being a functional structure for faculty as a future event space. This gazebo replaces an existing one, revitalizing the space in conjunction with the arboretum’s planned growth over the coming years. There were many challenges throughout the construction process including permitting issues, design complexities, and coordination with architects and engineers. Persevering through these taught a great deal about project management, the construction industry, and the importance of teamwork. The structure is a total rejuvenation of the prior structure; it is now stronger, …

Developing A 4d Logistics Educational Module With Cmbuilder, Misael Gonzalez Ramirez

Construction Management

The increasing adoption of Building Information Modeling (BIM) in the construction industry has led to a need for its increasing integration into construction management education. While existing curriculum covers various BIM applications, there is often a gap in in-depth instruction on BIM-based site logistics planning. This paper details the development of an educational module designed to address this gap, focusing on 4D site logistics and phasing using cmBuilder. The module aims not only to provide students with technical software skills but also to equip them with practical skills in visualizing, planning, and communicating complex construction site logistics. Through a series …