A machine learning project for predicting bus delays across the Georgia Tech campus shuttle lines, based on historical GPS data, time-of-day, and route information.
Our goal is to develop an end-to-end system that accurately predicts bus arrival delays using machine learning models. This project aims to improve the commuter experience, enable better student planning, and provide data-driven insights into operational inefficiencies.
The project is divided into 3 separate subteams, each focused on a core component of our project:
Goal: Automate the data pipeline for collecting, processing, and storing GPS and transit data.
Responsibilities:
- Scrape real-time GPS data
- Clean and process collected data
- Set up PostgreSQL for structured data storage
Members:
- Roy Chung https://www.linkedin.com/in/royschung/
- Diana Garcia Perales https://www.linkedin.com/in/diana-garcia-perales-gatech/
- Savvy Dusad
- Jevon Isaac Twitty
Goal: Design, train, and evaluate predictive models for estimating bus delays.
Responsibilities:
Members:
Goal: Build an interactive frontend that displays delay predictions to users in a clear and intuitive way.
Responsibilites:
- Design UI mockups using Figma
- Build the web interface using React
- Integrate the backend with the other two subteams
Members:
- Hieu Nguyen https://www.linkedin.com/in/hieu-nguyen-px27/
- Sahil Hora