Dublin Bikes

A small everyday question: will there be a bike when I get there?

A map can show you where a bike station is. Making that information useful for a journey also means understanding availability over time.

This project collected three months of station occupancy data from JCDecaux and used a random forest model to explore predictions of bike availability. A web application brought the station information and predictions together.

The work connected data collection, a prediction model, and a practical interface: a complete path from a changing city to something a person could use.

Python / Flask / JavaScript / AWS EC2 / RDS