This project is a full-stack system for remotely controlling a fleet of small differential-drive robots through a web browser. A React dashboard backed by Three.js renders a live top-down map of the arena, showing each robot's position, heading, and sonar readings in real time over WebSockets. Multiple users can share control of the fleet through a first-come, first-served queue, while robots without an active operator explore the arena autonomously under server-side collision prediction and onboard ultrasonic obstacle detection. The system spans three layers: embedded Propeller Spin firmware running on each ActivityBot, a Python asyncio server managing fleet state and network connections, and a browser-based interface for visualization and control.

8:20am, Thursday, April 30, GMS 1096

This project’s goal is to use a machine learning approach to discover new relationships and accurately classify a prediction on whether or not a social media post will remain popular. It achieves this by observing to see if it has sustained growth and if it reaches the top amount of posts based on views. The project runs off of a flask app and uses python to visualize results.

8:45am, Thursday, April 30, GMS 1096

This immersive, spatial-computing application for the Apple Vision Pro allows homeowners to visualize interior and exterior remodeling choices on their own home in real-time. By enabling an interactive way of swapping materials, colors, and textures on walls under dynamic conditions, it provides a realistic experience that eliminates design uncertainty. This tool transforms expensive guesswork of home renovation into a confident decision-making process for homeowners.

9:10, Thursday, April 30, GMS 1096

This project explores the ripple effects of the supply chain through a live, interactive web dashboard that predicts the price of a cup of coffee. Powered by a linear regression model, the system allows users to input custom price shocks for raw ingredients like milk, sugar, and coco. By calculating these user inputs against historical data, the application visualizes the true economic impact on the consumer and highlights the surprisingly strong correlation between sugar and overall coffee prices.

8:30am, Saturday, May 2, GMS 1097

My project uses play-by-play data from the National Football League (NFL) to develop a play-level model predicting how down, distance, and field position influence the points scored on a drive. The model estimates points scored based on the result of the drive, averaging these values to get a value that estimate the expected points for a given play call. This value can then be used to compare different situations to determine what is the better play call to make for the specific situation, or figure out which teams are better at scoring in certain situations.

8:55am, Saturday, May 2, GMS 1097

This versatile application will enable a user to follow a sequence of clues, guided by GPS and beacons (a device that emits BLE). This application also has an admin application where a admin user can create and set up different games for a user to play on the main application.

9:20am, Saturday, May 2, GMS 1097

This project focuses on computational redistricting, using a recombination algorithm to generate and analyze alternative district maps. This process was used to generate thousands of possible district maps through simulation. The goal is to explore how different district maps impact fairness, population balance, and partisan outcomes.

10:00am, Saturday, May 2, GMS 1097

Uncertainty in the market deters many novice investors from dipping their feet into investing and the stock market. My tool sets out to help and guide these investors to make educated investments by leveraging historical data as well as making sense of the news around companies of their choice.

10:25am, Saturday, May 2, GMS 1097

Survival of the Prettiest is an interactive application used to reconstruct a user selected image using polygons of various colors and sizes via a genetic algorithm. The application allows the user to be in control, therefore it is designed to receive user input to run the algorithm. The genetic algorithm runs on a population of individuals - the objects containing the polygons, and improves the individuals in a population one generation at a time using selection, mutation, and crossover.

10:50am, Saturday, May 2, GMS 1097

This project involves developing a computer vision algorithm to convert 2D images of a physical chessboard taken from skewed angles into a standardized digital game state, and creating an application that automatically tracks the match from start to finish in live time. The application allows users to play naturally while the system intelligently filters out physical obstructions like human hands, continuously detecting legal piece movements through edge detections, these moves are then converted into a universally recognized format (PGN) for digital review and engine analysis.

11:30am, Saturday, May 2, GMS 1097

Major League Baseball pitchers attempt to "expand the strike zone" by varying factors such as their pitch location, speed, and sequence to get batters to swing at pitches that would otherwise be called balls. This analysis explores how and when pitchers are able to manipulate the strike zone to their advantage through an at-bat simulation.

11:55am, Saturday, May 2, GMS 1097

A study into using past ATP rankings data to predict future rankings. Employs factors such as player age to build predictions. Additionally looks to specifically identify players poised for a breakout season before it happens. Findings are pulled into an interactive tool for player view and comparison.

1:15pm, Saturday, May 2, GMS 1097

My project is a web application where people can upload photos from their trips and adventures, so they can see where they have gone on a map. User's can also see where other users have gone on their trips. The application allows for users to upload, edit, delete, and filter photos that they have uploaded.

1:40pm, Saturday, May 2, GMS 1097

This project examines how specific housing characteristics influence property prices across Wisconsin. Using statewide Zillow housing data that includes geographic location and home features, the analysis estimates the value added by characteristics such as additional bedrooms and bathrooms. To estimate housing prices, a non-parametric model is used, allowing the data to speak for itself while generating accurate price predictions. This approach provides insight into how home features contribute to value across Wisconsin housing markets and relates the significance of certain features to how they affect housing prices.

2:15pm, Saturday, May 2, GMS 1097

Eye-Spy is a hidden object game played with your eyes. Utilizing the Tobii Pro Spark eye tracker, the user's vision is tracked to determine if an object is found. The user can then replay their vision path as well as other's. An additional admin application allows for custom pages to then be played.

2:40pm, Saturday, May 2, GMS 1097

Bit manipulation is the process of changing information on the binary level. My program uses this technique to hide image information and obscure it in a similar way to encryption. There are two techniques one may use to encode and decode in this program: hiding an image within another image using pixel binary data, and fracturing an image into shares which look like fuzzy, incomprehensible noise. The decoder reverses these processes to reveal the secret images.

TBA