CityOS

Project Information

Documentation

Weekly Progress

Week 1

• Met with Dr. Ortiz to define project scope.
• Began researching and learning about data regression models.
• Reading documentation on similar works.

Week 2

• Drafted the detailed system design document.
• Obtained video feed of the parking lot and applied YOLOv8 for object detection.
• Outlined the system architecture, including the Data Logger, ML Model Trainer, and Prediction API.

Week 3

• Implemented and tested three initial ML models: Linear Regression, Naive Bayes, and Gradient Boosting.
• Achieved initial prediction results on sample data.
• Began learning about Decision Tree and Random Forest algorithms for future implementation.

Week 4

• Implemented and tested Decision Tree and Random Forest ML models.
• Created the website, which holds documentation and the web app.
• Continued with data collection and aggregation for ML models.

Week 5

• Created and filled the design documentation section of the website.
• Designed a mock-up layout of the web app with components.
• Collected and imported more data, as well as correcting some overfitting with the model
• Ran RandomCV and GridsearchCV across multiple folds for optimization and cross-validation
• Generated occupancy histograms per day to visualize trends in each lot

Week 6

• Created 3D model assets of vehicles for the web app.
• Generated occupancy graphs from the machine learning model.
• Continued collecting more data to train the model.

Week 7

• Implemented a finished rough layout of the parking web app and connected it to the website.
• Analyzed graphs outputted by the Random Forest's prediction and the actual camera detection of that day.
• Conceptualized an expansion of the project utilizing more of the parking lot, which requires a new detection model not based on bounding boxes or points.

Week 8

Week 9

Week 10

Other Information