HackMIT 2021 Hurricane Damage Detector

Hack MIT 2021 Project Repository: Hurricane Damage Detection Web App

How to run

Note: To reduce install time, run in conda env

cd python &&  pip3 install -r requirements.txt && python3 main.py

then in seperate terminal

cd web && npm i && npm run start

Team

Jacob Duncan - Salisbury University
Justin Ventura - Salisbury University
Blaine Mason - Salisbury University
Jenna Bustami - University of California Berkley

Inspiration:

With a majority of the team being from Salisbury, Maryland we encouter many hurricanes each year. Luckily, what makes impact near us is rarely fatal or destructive, but with a rapidly changing climate we cannot guarantee safety.

Just earlier this August, hurricane Ida tragically made landfall in Louisiana leaving many without housing, electricity, and in some cases, their lives. Information about hurricane evacuation and/or preparing for a hurricane is not very easily accessible, and the motivation behind this web application is to not only raise awareness, but to help those potentially in danger be prepared to take action during a hurricane before it's too late.

What it does:

Hurricane Damage Detection provides resources to evacuate/prepare for a hurricane, historical information about previous hurricanes, and a live feed for storms. Currently, we are able to provide information about severity of past storms(1980s-Present), but unable to provide live danger measurements. The only thing missing is the function to compute this measurement. We attempted, but with each variable in a tropical cyclone it is challenging to manage the weights. We would like to possibly investigate doing this in a research project.

What's next for Hurricane Damage Detector

In the future we hope to add the following features:

The ability to get evactuation routes based on current location.
Live damage updates, for instance when the storm is 200km out allow nearby counties to know how in danger they are.
- We were close to this, but ran out of time to test all cases.
More information provided after the questionnaire is filled out.
The ability to download the .csv of past storm.
Provide tools like nearby Zipcars to assist in evacuation.

What we learned

How to develop a fullstack application.
Working as a team to provide efficient code to who needs it.
Sleep is required for the best work to be done.
Communication helps avoid unnecessary confusion.

Technicals:

Tech Stack: JavaScript, React, TailwindCSS, Formik, Python, Flask
API: Google Maps API
Data: National Hurricane Center Dataset

jacobduncan00/hackmit2021