- The recent bush fires in Australia
- The increasing risk of other extreme climate events due to climate change
- The need for better data gathering to enable researchers to make better predictions
- Increase fire detection infrastructure, provide platform to make information accessible to researchers and citizens
- Providing a way for everyone in local communities to get involved
What it does
Stage 1: Built low-cost Arduino sensors to detect temperature and humidity levels. Stage 2: Data uploaded via Wi-Fi connection (Internet of Things) to Google Cloud app engine Stage 3: Integration of data streams into predictive model for fire risk Stage 4: Heatmap data visualisation of fire risk in a web app hosted on Microsoft Azure platform
Challenges I ran into
Pivot from initial idea of more data analysis/machine learning-based project after we realised that there was more potential in IoT applications to solving the problem. SQL. Ill-advised attempt to get stronger signals from our temperature and humidity sensors by placing directly above a cup of boiling water led to frying the breadboard and emergency rebuild of one of our sensors.
Accomplishments that I'm proud of
Implemented REST API for the first time, First time working with hardware. First time creating a web app, and hosting it on the cloud as an API First time working with cloud platform.
What's next for Save the Burning Forest
Drone flights over risky areas - providing high definition local information. With more data types, and platform that allows anyone to contribute. Inviting citizen scientists to contribute predictive models, that can be layered onto our website.