Inspiration
We wanted to expand on the initial winning idea of a ML classifier beyond its 3 basic classification categories of low, medium and high, to also incorpate mobility constraints that make it hard to leave an area. Moreover, it highlights at-risk areas that the general public should avoid, and vulnerable areas for first responders to render aid and evacuation.
What it does
A real-time ML risk heatmap of Singapore that combines live rainfall APIs with areas of higher mobility constraints like traffic congestions, hospitals, childcare and elderly centres. This highlights areas with potential evacuation difficulties, helps in the deployment of first responders to areas more in need and also helps identify areas at risk of flash floods.
It classifies zones into 4 main categories: 0 - Normal; 1 - Monitoring Alert; 2 - Mobility Warning; 3 - Critical Flash Flood Alert.
How we built it
Challenges we ran into
Accomplishments that we're proud of
What we learned
What's next for HydroFlash Move
Built With
- apis
- databricks
- datasets
- python