SpaceFinder is a web application that predicts and identifies available parking spots using real-time occupancy data.
Project Highlights
- Built an IoT system using 9 sensors and 2 actuators integrated with Firebase real-time database for collecting parking space occupancy data.
- Developed a machine learning model using Gradient Boosting algorithm to predict parking availability across multiple spaces, achieving over 83% accuracy.
- Created a Flask web application that displays predicted availability status for parking spaces to drivers.
- Designed system architecture and data pipelines for ingesting sensor data, training ML models, and serving predictions to web app.
- Implemented model retraining and result monitoring to maintain prediction accuracy over time as new data arrives.
- Gained valuable experience in developing an end-to-end IoT and machine learning system.
Working with 11-person in team this project gave me hands-on experience in working with IoT devices, setting up cloud data infrastructure, training supervised learning models, and building web applications. I enjoyed the opportunity to build a solution that combines my interests in hardware, software and ML.
