Real-time On-Street Parking Recommendation Engine
Winner of Sustainable Award Coding Fest 2021 @USYD
Introduction
WHAT IF you know in advance where you can park off-street near your destination?
WHAT IF we guarantee you an off-street parking only 3 minutes walk from your destination?
WHAT IF Google Map directs you to the best parking when you are near your destination?
With real-time parking sensors data released by the City of Melbourne (link), we created a recommendation system to help drivers navigate to the most satisfying off-street parking bay. Using deep learning model, our system not only considers distances, cost and restriction but also predicts how likely you can park at an off-street bay near your destination. We hope to start a Proof-of-concept that can improve traffic in busy suburbs and reduce air pollution by directing vehicles.
Requirements
- Python 2.7+
- Flask 2.0+
- MongoDB 4.0
How do we rank a parking spot?
- Close to the destination
- Better price
- Free of restriction (you can park longer)
- High probability to park the car (work in progress)
Future Work
- Continue to improve UI
- Add car park commercial car park information
- Online allocation to prevent multiple drivers competing the same parking bay