Engineering Commons^™

Ai-Based Hazard Detection For Railway Crossings, Darren Espinoza, Gasser G. Ali, Constantine Tarawneh

Mechanical Engineering Faculty Publications

Grade crossings are critical elements of the railway infrastructure due to the potential risk of vehicle collisions with trains. According to the National Highway Traffic Safety Administration, there were more than 1,600 vehicle-train, and 500 human-train collisions in 2020. Researchers, transportation organizations, and government bodies are constantly exploring practices and technologies to improve safety at crossings. Examples of safety standards include sensors, motion detectors, depth cameras, and many other innovative technologies. The goal of this paper is to investigate the applications of computer vision using Artificial Intelligence (AI) deep learning models to enhance railway safety. Deep learning models can provide …

Modified Method For Assessing The Required Expenditures And Estimated Time Of Hard Coal Mine Liquidation, Andrzej Chmiela, Małgorzata Wysocka, Adam Smoliński

Journal of Sustainable Mining

The restructuring of hard coal mining requires significant budgetary expenditures. A comprehensive scientific approach may facilitate the rationalisation and minimisation of mine closure costs. This study proposes a method for the preliminary estimation of the costs and time required for the potential liquidation of a hard coal mine. In addition to a literature review, a statistical analysis and a case study, personal interviews were conducted with individuals with direct management over the restructuring, reclamation and liquidation processes pertaining to mines undergoing closure.

The assessment method is based on an analysis of the mine liquidation costs, divided into successive years of …

Deep Learning Based Single Image Super-Resolution, Samuel Smith

Computer Science and Engineering Senior Theses

Single image super-resolution (SR) involves taking a given low-resolution (LR) image and generating a corresponding high-resolution (HR) image. This is a core task in the computer vision field due to its multitude of applications ranging from helping current-day issues of storage and transfer of data to restoration of low-resolution images. The current strategies, however, struggle to reach the quality needed for their widespread use and are often too resource-intensive for the average consumer. While other lightweight SR techniques exist with different techniques, like Cascading Residual Networks [3], their success comes at the cost of expensive technology inaccessible to most situations. …

Distant Horizon: Exploring Human-Ai Interaction Through Video Games, Gabe Labadie, Dalia Suszko, Max White

Computer Science and Engineering Senior Theses

As the field of Artificial Intelligence (AI) continues to grow and become more and more integrated into our everyday lives, it has begun to raise ethical concerns surrounding authority, autonomy, and responsibility. Researchers and consumers alike have begun to wonder how much trust we are willing to place in a non-human decision maker, especially if an Artificial General Intelligence (AGI) capable of surpassing human cognitive capabilities is ever developed. If we are willing to let an AI write essays for us, are we willing to let it manage a business? A government agency? What about a nuclear reactor?

Our project …

Daily Digest: A News Aggregation Site, Justin Wang, Jack Maguin, George Orloff, Justin Groves

Computer Science and Engineering Senior Theses

A staggering amount of news articles are uploaded every day – approximately 5,000 in the United States alone [1]. That volume of information causes difficulty for many people who try to stay up-to-date with current events. The number of articles and the multitude of sources that they come from can feel overwhelming. In our project, we attempt to tackle this issue. We use web scraping to collect a dataset of news articles and combine it with a Large Language Model (LLM) capable of processing those articles to generate news summaries for the user. The user interacts with the program through …

General Purpose Tuning Data Visualization, Chris Augustine, Francisco Salinas, Aakash Shetty

Computer Science and Engineering Senior Theses

This project centers on the visualization of High-Performance Computing (HPC) data obtained from the GPTune website. GPTune serves as a valuable resource for HPC experiments, providing a wealth of performance data from tuning studies and optimization tasks. Our objective is to develop an advanced data visualization framework tailored to GPTune’s datasets. Utilizing state-of-the-art visualization techniques, we aim to create an interactive platform that allows users to explore, analyze, and derive insights from the diverse tuning experiments conducted on HPC systems. The visualization tool will facilitate the identification of optimal configurations, performance trends, and patterns within GPTune data, empowering researchers and …

Applied Auto-Tuning On Lora Hyperparameters, Darren Inouye, Lucas Lindo, Robin Lee, Edmund Allen

Computer Science and Engineering Senior Theses

This senior design project explores the application of Bayesian optimization-based auto-tuning techniques on the low-rank adaptation (LoRA) fine-tuning of large language models (LLMs), demonstrating how fine-tuning methods can reduce training times and costs, albeit with a slight trade-off in accuracy. However, little is known about the optimal hyperparameters on LoRA and its variants for those methods. This project addresses this lack of knowledge by analyzing data gathered from auto-tuning LoRA hyperparameters to determine the most optimal parameter configurations for a model’s accuracy and training efficiency.

The team has implemented a pipeline utilizing many different technologies. The main technology driving the …

9-Axis Motion Tracking To Aid Therapeutic Recovery Via Visualization, Analysis And Progress Monitoring, Megan Wiser, Liam A'Hearn, Christopher Tamayo, Liam Kelly

Computer Science and Engineering Senior Theses

This paper presents an innovative approach to enhance at-home physical therapy exercises through the development of a wearable motion tracking system. The proposed system utilizes motion tracking bands worn by patients during exercises, specifically focusing on a squat jump for the initial phase of the project. The bands, placed around the ankle and knee, monitor the alignment of the user's motion and transmit data via Bluetooth Low Energy (BLE) to a dedicated webpage. This webpage integrates real-time data analysis, offering immediate feedback to users, enabling them to monitor their form and track progress over time. The collected data is stored …

Federated Learning Based Autoencoder Ensemble System For Malware Detection On Internet Of Things Devices, Steven Edward Arroyo

Theses and Dissertations

New technologies are being introduced at a rate faster than ever before and smaller in size. Due to the size of these devices, security is often difficult to implement. The existing solution is a firewall-segmented “IoT Network” that only limits the effect of these infected devices on other parts of the network. We propose a lightweight unsupervised hybrid-cloud ensemble anomaly detection system for malware detection. We perform transfer learning using a generalized model trained on multiple IoT device sources to learn network traffic on new devices with minimal computational resources. We further extend our proposed system to utilize federated learning …

How Ai And Digitalization Can Help To Improve Managing Projects In Lean Manufacturing & Supply Chain Environment, Abhijeet Hange

Harrisburg University Dissertations and Theses

This research examines how digitalization and machine learning (ML) revolutionize project management in supply chain and lean manufacturing settings. To improve the efficiency and caliber of their operational procedures, Acuity Brands, Amazon, Walmart, Apple, and other medium-sized businesses—which are at the center of today's competitive markets—are resorting to automation and digitalization. With an emphasis on the fusion of digital and artificial intelligence (AI), the paper explores how successful transformation initiatives are essential to the automation and digitization of manufacturing and distribution hubs. The study uses quantitative methods approach to evaluate the benefits and drawbacks of these technology integrations by integrating …

Pixel-Mps: Stochastic Embedding And Density-Based Clustering Of Image Patterns For Pixel-Based Multiple-Point Geostatistical Simulation, Adel Asadi, Snehamoy Chatterjee

Michigan Tech Publications

Multiple-point geostatistics (MPS) is an established tool for the uncertainty quantification of Earth systems modeling, particularly when dealing with the complexity and heterogeneity of geological data. This study presents a novel pixel-based MPS method for modeling spatial data using advanced machine-learning algorithms. Pixel-based multiple-point simulation implies the sequential modeling of individual points on the simulation grid, one at a time, by borrowing spatial information from the training image and honoring the conditioning data points. The developed methodology is based on the mapping of the training image patterns database using the t-Distributed Stochastic Neighbor Embedding (t-SNE) algorithm for dimensionality reduction, and …

Chemical Herding As A Multiplicative Factor For Top-Down Manipulation Of Colloids, Mark N. Mcdonald, Douglas R. Tree, Cameron K. Peterson

Faculty Publications

Colloidal particles can create reconfigurable nanomaterials, with applications such as color-changing, self-repairing, and self-regulating materials and reconfigurable drug delivery systems. However, top-down methods for manipulating colloids are limited in the scale they can control. We consider here a new method for using chemical reactions to multiply the effects of existing top-down colloidal manipulation methods to arrange large numbers of colloids with single-particle precision, which we refer to as chemical herding. Using simulation-based methods, we show that if a set of chemically active colloids (herders) can be steered using external forces (i.e., electrophoretic, dielectrophoretic, magnetic, or optical forces), then a larger …

A Comprehensive Study On The Impact Of Human Hair Fiber And Millet Husk Ash On Concrete Properties: Response Surface Modeling And Optimization, Naraindas Bheel, Muhammad Alamgeer Shams, Samiullah Sohu, Abdul Salam Buller, Taoufik Najeh, Fouad Ismail Ismail, Omrane Benjeddou

Department of Civil and Environmental Engineering: Faculty Publications

Revolutionizing construction, the concrete blend seamlessly integrates human hair (HH) fibers and millet husk ash (MHA) as a sustainable alternative. By repurposing human hair for enhanced tensile strength and utilizing millet husk ash to replace sand, these materials not only reduce waste but also create a durable, eco-friendly solution. This groundbreaking methodology not only adheres to established structural criteria but also advances the concepts of the circular economy, representing a significant advancement towards environmentally sustainable and resilient building practices. The main purpose of the research is to investigate the fresh and mechanical characteristics of concrete blended with 10–40% MHA as …

Emergent Magnetism And Hyperthermia In Phase- And Size-Tunable Iron Oxide Nanostructures, K Mudiyanselage Tharindu Supun Bandara Attanayake

USF Tampa Graduate Theses and Dissertations

Iron oxide nanoparticles (IONPs) hold immense potential across diverse fields, from spintronics, magnetic hyperthermia to drug delivery, biodetection, and magnetic resonance imaging. Unlocking these applications hinges on our ability to tailor the magnetic properties of IONPs, and this dissertation presents novel and powerful approaches for enhancing magnetic and hyperthermic responses of multiphase iron oxide nanostructures. This is achieved by manipulating their phase volume fraction, size, and shape with a focus on the magnetic tunability of nanostructures via strategic control over structural and environmental parameters. The aforementioned has been achieved by the following three key approaches: (i) phase-tunability of iron oxide …

Physical Effects On The Worst-Case Delay Analysis And Signal Integrity Of Buses And Spirals, Mahmoud Mahany

Theses and Dissertations

Physical effects have a significant impact on the IC design which will be investigated in this thesis. Moving toward advanced technology nodes, magnetic effects become more dominant than capacitive effects. As the dimensions of the devices go down and the interconnect manipulates the circuit behavior more and more. Cross talking and voltage drops are affecting the design heavily, however - going to the full electromagnetic point of view - current return path (CRP) adds significant parasitics to the performance of the chip. Neglecting the CRP gives wrong intuition and simulation of the designs, especially that the environment and surroundings can …

Assessment For Sustainable Transportation Systems In Egypt, Hagar Mohamed Fares

Theses and Dissertations

The assessment of a sustainable transportation system poses a significant challenge for local governments, as it dramatically influences the promotion of environmentally friendly modes of transportation with the objective of safeguarding the public's welfare and health in their daily activities. The integration of such assessment should be implemented from the initial stages of system planning to its operational and maintenance phases, particularly in light of the increasing number of urban development activities nationwide. In order to facilitate decision-making processes, it is imperative to establish an efficient rating system that incorporates cultural and social constraints and comprehensively considers the three dimensions …

A Machine Learning Framework For Predicting Fabrication Hours For Industrial Steel Structure Projects, Dalia Ibrahim

Theses and Dissertations

Construction projects are considered high risk projects especially due to their required large capital making them require extreme attention in estimation as overestimating a project will lead to losing bids and underestimating them will lead to incurring more costs than budgeted resulting in losses. However, estimators are often faced with very tight timelines to finish their estimates leading them to primarily rely on their experience disregarding some crucial factors resulting in inaccurate estimates. In the steel structures industry, the steel fabrication phase accounts for 30 to 40% of the overall project cost; in addition, the steel industry is labor driven; …

Electrochemical Identification Of Metal Chlorides In Eutectic Licl-Kcl Without Prior Knowledge Of Analyte Identities, Tyler Williams, Jason Torrie, Mark Schvaneveldt, Ranon Fuller, Greg Chipman, Devin Rappleye

Faculty Publications

The identities of unknown analytes within four eutectic LiCl-KCl melts were determined using electrochemical methods, simulating the uncertainty of electrochemically probing an electrorefiner salt bath or molten salt nuclear reactor. With a variety of electrochemical methods (e.g. cyclic voltammetry, chronopotentiometry, and square-wave voltammetry), and electroanalytical techniques (e.g. semi-differentiation), every analyte was positively identified, although one false positive occurred because of an unexpected chemical interaction. This study highlights some remaining challenges for the use of electrochemical sensors in nuclear material control and accountability in molten salts: (1) quantification of analytes without the use of calibration curves (e.g. error in property values, …

An Efficient Method For Recognition Of Human-In-Motion Action Based On Foot-Lift Features, Khin Cho Tuna, Hla Myo Tunb

ASEAN Journal on Science and Technology for Development

In the age of Industry 4.0, recognition of human moving actions becomes an essential element in smart surveillance systems at public places. The requirement of a large number of frames, view dependency and the requirement of big database are still challenging issues for human action recognition in real-time applications. This study proposes an efficient method for basic moving actions, “Walking” and “Running”, based on foot-lift features aiming at the recognition of ongoing moving action by reducing the number of frames required. The plane of moving path and foot-lift are estimated during the action by means of BLOB analysis. The performance …

Hiv3: An Efficient Beehive Monitoring System, Jack Ursillo, Aneal Kuverji, Connor Merhab, Anshuman Sahu

Computer Science and Engineering Senior Theses

Beehive monitoring plays a major role in ensuring the health of beehives by checking for overpopulation or underpopulation within a hive. Beehive monitoring provides beekeepers with the opportunity to take action and save the hive before the problem becomes irreversible. Most solutions are too expensive for everyday beekeepers and lack elements of sustainability, making it impractical for small scale beekeepers. In this thesis, we propose a solution to this problem, demonstrating its sustainability and user-friendliness, which enables us to effectively reach a larger consumer market. We support these claims through the use of sustainable systems such as using a solar …

Enhancing Vqgan-Based Model With Connext For Blind Super-Resolution, Lebin Zhou

Computer Science and Engineering Master's Theses

This thesis presents a novel super-resolution model based on Vector Quantized Generative Adversarial Network (VQGAN) to enhance image resolution. Inspired by recent advancements in the field of image reconstruction, we apply VQGAN to the super-resolution task, leveraging its powerful generative capabilities to produce higher quality high-resolution images.

Building on the VQGAN framework, we propose an improved architecture that incorporates an additional ConvNeXt feature extractor based on Convolutional Neural Networks (CNN) to effectively capture and refine features from low-resolution images. To further enhance model performance, we implemented various strategies to optimize the utilization of the codebook, including capacity optimization, improved initialization, …

Finding The Shortest Path Using Dijkstra’S Algorithm, Orit D. Gruber, Deborah Sturm

Open Educational Resources

This lab experiment explores an algorithm which is used to find the shortest path between two or more locations. After completing the lab, you will be able to answer the following questions in the final lab report:

1. What is an Algorithm?
2. What is a Graph ?
3. What is the purpose and operation of Dijkstra’s Algorithm ?

Transforming Orthopedic Surgery: Autonomous Image-Guided Techniques For Femur Fracture Robotic Surgery, Marzieh Sadat Saeedi-Hosseiny

Theses and Dissertations

Long-bone fractures, particularly femur fractures, are prevalent and necessitate surgical intervention due to traumatic forces. Manual reduction procedures pose challenges, including excessive traction forces and reliance on limited X-ray imaging, resulting in malalignment-related complications and repeated surgeries. Current methods for femur fracture reduction are prone to inaccuracies, prolonged surgical times, and postoperative complications. This research introduces image-guided methods to automate a surgical robotic system, Robossis, to enhance femur alignment accuracy, reduce procedure times, and improve patient outcomes. To address these challenges, a marker-based approach utilizing a few X-ray images is developed to detect the spatial relative position of the femur …

Design And Fabrication Of Passive Microfluidic Devices For Enhancing Membrane-Based Biosensors, Tanner N. Wells

Theses and Dissertations

Microfluidic devices and systems have demonstrated wide applicability in sample processing and analysis. The ability to integrate sample processing and analysis into a single platform is a key benefit of microfluidic technologies. This dissertation describes the design and fabrication two types of microfluidic devices developed to enhance membrane-based biosensors. First is a device for collecting particles into a highly confined region, locally enhancing their concentration. Second is a device for size-based particle separation capable of separating differently sized particles into separate streams. Both types of devices are fabricated with a thin top layer, providing a platform for synthetic membrane-based technologies. …

The Design, Prototyping, And Validation Of A New Wearable Sensor System For Monitoring Lumbar Spinal Motion In Daily Activities, Brianna Bischoff

Theses and Dissertations

Lower back pain is a widespread problem affecting millions worldwide, because understanding its development and effective treatment remains challenging. Current treatment success is often evaluated using patient-reported outcomes, which tend to be qualitative and subjective in nature, making objective success measurement difficult. Wearable sensors can provide quantitative measurements, thereby helping physicians improve care for countless individuals around the world. These sensors also have the potential to provide longitudinal data on daily motion patterns, aiding in monitoring the progress of treatment plans for lower back pain. In this work it was hypothesized that a new wearable sensor garment that makes use …

Nucleation Control System And Method Leading To Enhanced Boiling Based On Electric Cooling, Amitabh Narain, Soroush Sepahyar, Divya Kamlesh Pandya, Vibhu Vivek

Michigan Tech Patents

A cooling module for an electric devise includes a body having formed therein a plurality of channels, a micro-structured boiling surface, a piezoelectric transducer, an inlet header, and an outlet header. Each channel of the plurality of channels is defined by a first channel surface and opposing lateral channel surfaces cooperatively defining a rectangular cross section normal to a channel axis The micro-structured boiling surface is positioned adjacent the first channel surface of each channel. The piezoelectric transducer is in acoustic communication with one of the opposing lateral channel surfaces of each channel and configured to direct acoustic waves on …

Ultra Low-Power, High-Performance Presence Detection System, Dante Bajarias, Cristian Medal, Jashan Kaeley

Computer Science and Engineering Senior Theses

The proliferation of Internet of Things (IoT) devices emphasizes a greater connection between humans and smart technology. From computer peripherals, personal electronics, and domestic appliances to building access management, healthcare, and security systems, modern applications are generating as much data or more than they are receiving from other sources. Human presence is a type of data that is becoming a compelling requirement for a plethora of today’s applications that must ensure seamless interactions of users with the systems and their associated services. The ultra low-power, high performance presence detection system is an accelerometer-based system capable of person detection, people counting, …

Manipulating Colloidal Particles Using Chemical Gradients And Top-Down Control, Mark Nichols Mcdonald

Theses and Dissertations

Colloidal particles provide the ideal building blocks for the next generation of microdevices, such as advanced sensors and precision drug delivery systems. However, many such applications require the use of top-down (i.e. humanly controllable) forces to manipulate colloidal particles with single-particle precision, and current methods can only achieve such precision for small numbers of particles at a time. To address this challenge, we propose using chemical forces in combination with existing top-down techniques to enable the control of larger numbers of particles simultaneously. Controlling colloids using chemical reactions is a novel technique not typically utilized. Due to its distinct difference …

Edge-Connected Microcontroller Security, Divya Syal, Gavin Ryder, Neena Ekanathan

Computer Science and Engineering Senior Theses

With a wide range of applications and the rise of cyber attacks, securing microcontrollers has become imperative; however, ensuring microcontroller performance is also crucial given how interconnected today’s systems are. This project examines the security and performance of next-generation microcontroller units leveraging new security solutions for IoT edge applications. By benchmarking these MCUs against key performance metrics, their viability will be assessed to facilitate the widespread adoption of this latest firmware.

Our research focuses on profiling the power consumption and performance of the new STM32H573 and the integrated Secure Manager, a new technology from ST Microelectronics that allows for privileged …

Autonomous Microgrid System, Xavier Kuehn, Brian Xiong

Computer Science and Engineering Senior Theses

Microgrids have made a revolutionary change in the realm of energy distribution due to the features that they offer, including localized, resilient, and sustainable energy solutions. Operating renewable resources in a microgrid while maintaining generation-load balance and acceptable voltage-frequency limits has been an open research problem. This thesis presents smart python agents for microgrid systems to automate the operations and control of microgrid renewable resources in an effort to provide resilient solutions to the intermittence issues that could potentially arise within the microgrid energy system. The smart agents operate the microgrids by not only integrating the use of renewable energy …