TUM.ai_Respons[Ai]ble

• 

  Risk assessment for landslides

Inspiration

Natural disasters cause financial losses into the millions, loss of lives and livelihoods. Thus, it is essential to be able to estimate the risk of a natural disaster happening in order to prepare and react in a timely and appropriate manner. We chose landslides as they do not get as much awareness as e.g. floods, even though they still pose a significant threat.

What it does

Our web application assesses the risk for landslides given the coordinates and the time of year. Currently, it's limited to the West coast of the USA but can easily be extended to additional regions depending on the availability of data.

How we built it

We used the Global Landslide Catalog provided by NASA as a data source. Kernel density estimation served as a model architecture which was implemented in Python (scikit-learn). For the front-end, we used the low code platform bubble.io.

Built With

• bubble.io
• global-landslide-catalog
• python
• scikit-learn