Using Machine Learning to Help Prepare for Wildfires
Learn More: https://devpost.com/software/wildfiresai
Demo: https://youtu.be/JjXK1m8niyg
4x hackathon award winner.
- Honorable Mention at 2021 Congressional App Challenge
- First Place Overall at Epsilon Hacks II
- Best Execution at OwlHack 2021
- Best Environment Hack at Citro Hacks
Unlike other natural disasters such as hurricanes and floods, wildfires are not constrained to a certain time period during the year, making them unpredictable and unfortunately harder to combat. This is largely because 80% of wildfires are created by human error, such as through unattended campfires and discarded cigarettes. Because of the vast expanse of most forests, firefighters are often unaware of wildfires in their early stages. Due to the lack of fire department personnel, the demand for automated software to prevent wildfires from fatally progressing is increasing. While there are many existing applications that alert civilians about the presence of a wildfire, most lack the ability to provide assistance to governmental authorities in charge of wildfire management.
You can view the hosted version of our website here: wildfires.ml. Please note that some functions may not be fully operational due to memory allocation and storage issues.
Clone the repo to your system by running the following command:
git clone https://github.com/CMEONE/WildfiresAI.git
cd WildfiresAI
In order to install all dependencies, run pip install -r requirements.txt
in your terminal or command prompt.
To run the code, open a terminal and navigate to the root directory of this repository. Then, run the following bash commands:
export FLASK_ENV=development
export FLASK_APP=app.py
flask run
As import errors arise, you may need to install more modules by using pip install MODULE_NAME_HERE
.
Finally, once the program runs without errors, navigate to http://127.0.0.1:5000/
in your browser.