Senior Design

Team Name

Card Counters

Team Abstract

This project examines the role of artificial intelligence in modern software engineering through the AI-assisted full-stack development of an intelligent blackjack tutoring system. The research investigates how AI tools, particularly large language models, can support software development while also serving as embedded components within interactive applications. Using an AI-augmented development workflow, a web-based blackjack tutoring platform will be created that analyzes user gameplay, evaluates decisions against optimal basic strategy, and provides real-time explanations based on probability and expected value. Throughout development, AI tools will assist with tasks such as code generation, debugging, and design support, while a structured development log documents the extent and effectiveness of AI contributions. The project follows a design-and-evaluate methodology to assess the impact of AI assistance on development efficiency, code quality, and system design decisions. Findings from this study aim to provide practical insights and best practices for integrating AI into full-stack software development workflows.

Team Name

Robotics Research Group

Team Abstract

Earlier, most of the robotics research was focused on augmenting individual robot capabilities. However, later, the efficiency demonstrated by multi-robot systems in collaborative approaches that expedite task completion has catalyzed increased scholarly attention towards them. Our current research explores how multi-robots periodically reconfigure themselves to address human requests in temporally constrained situations and examines the effect on overall task completion on the Lunar surface. The researchers are also examining how robots behave when handling priority-driven and urgent and non-urgent human requests and will analyze how this affects coordination among multiple robots on the Lunar surface.

Team Name

Robotics Research Group

Team Abstract

Sending several robots to work together on the Moon is no easy feat: the terrain is rough, distances are large, and mistakes are expensive. Before building hardware, it helps to practice in a computer simulation where we can try ideas safely and repeat them many times. Our guiding question is: How can Gazebo be used to design and model multi-robot navigation on the lunar surface?

We use Gazebo Sim Harmonic to run four, 4-wheeled robots on a large, rough terrain that acts as a stand-in for a lunar surface. A key part of our setup is that a person can type in map coordinates and a robot is given a path to explore the region between where it started and that goal. The software breaks that region into smaller sections, drives toward each corner in order, and can return home after the goal is reached or after a time limit. This allows us to examine how navigation performs when the destination is selected dynamically rather than predetermined. Additionally, robots can respond to human commands , with different temporal complexity. In this work, combined simulation and user-defined goals enable us to rehearse multi-robot missions at designated surface locations prior to any hardware deployment.

Team Name

Calorie Burners

Team Slogan

“The only bad workout is the one that didn’t happen.”

Team Abstract

This project is meant to give anyone an opportunity to start or continue their workout journey in the gym. The main objective with this project is for everyone to achieve their goal with their body, whether that’s toning your body, losing weight, or even just staying consistent. AI gym Buddy gives you a generated workout plan that focusses on what you want to achieve. AI gym buddy also gives a calendar to keep track of your workout progress, and a workout summary tab to keep track of what you have done.

Team Name

AgriVision

Team Abstract

This senior design project explores how computer vision models can be made more transparent and trustworthy. The goal is not only to build models that can detect weeds and plant diseases but also to make their decision-making process visible. Unlike prior agricultural computer vision work that focused mainly on raw performance, this project emphasizes explainability, a pressing issue in trustworthy AI. Agriculture provides a practical and meaningful application area, while the broader research focus is on explainability; helping people understand why a model made a particular decision.

Team Name

Data mine for Image Recognition

Team Abstract

This project uses AI image detection and model training to accurately identify license plate characters based on the data set of vehicle images provided for future use on toll roads. The goal was to develop a fast and reliable system that can efficiently preprocess, and clarify images, detect license plates, recognize plate numbers, and improve overall accuracy through training and optimization.

Team Name

Prediction of Gini Coefficient

Team Abstract

The project applied Deep Learning technology for creating a model to predict the most important indicator of income inequality, Gini coefficient in the future, based on the historical relevant data, so that observe the ever widening gap between the rich and the poor. It also found the major elements that governed this widening behavior through analyzing the impacts of the related attributes. This study obtained and analyzed substantial amount of data, which contain information on the income, expenses, and the financial footprints of families in the United States, to draw empirical conclusions. The results may help the public and the economic research society for their decision making. Using Deep Learning algorithms, the data analysis was far more efficient, and by generating the Deep Learning multi-layer Neural Networks, the prediction was quite accurate.

Team Name

Bright Idea Games

Team Abstract

This project provides hands-on experience with the Unity game engine through the development of a third-person hero shooter. The game features local multiplayer and a diverse selection of weapons, each offering unique abilities that players can switch between during gameplay. The project establishes a solid and functional foundation, designed with the flexibility to support future expansion and refinement.