Inspiration
Natural disasters are an inevitable part of life but people receiving disaster aid doesn’t have to be. Some disaster relief organizations and plans require credentials before aid is sent, but there are plenty of nonprofits and volunteer supplied organizations that do not. As soon as an event happens they begin to gather supplies and volunteer personnel and move out. But what if they didn’t have to wait until after the event has happened to start gathering volunteers. What if they know as soon as disaster data is collected for storm surges, floods, hurricanes, tornadoes,etc. that there would need to be personnel on-site as soon as the event is over.
What it does
Our project is a WPF application that interacts with an R script and a UiPath file. The UiPath file web scrapes a website so that we were able to get the information about the previous natural disasters. The UiPath scraped over 61,000 data entries and imported it into a CSV. The WPF application takes the information from the CSV and adds it to the Firebase Realtime database. It also allowed users to create a Disaster Relief Form and submit it to the database. The R script went through the data using Machine Learning to find out if the the created forms will receive FEMA aid based on the criteria from past disaster information taken from the website.
How I built it
We used Firebase to build the database, UiPath to import the open source dataset to the database and to build a webscraper to pull data from natural disaster sites, R to code the machine learning algorithm and C++ to take data from the database and modify it for useful information.
Challenges I ran into
- Finding a meaningful way to use the open source data that we found about natural disasters
- Using UiPath, specifically how to use it and connecting it to the Firebase
- Learning how to make a machine learning algorithm and learning R
- Trying to connect everything together after completion
- Overall, just trying to accomplish our goal withing the time-frame given ## Accomplishments that I'm proud of
What I learned
Bucci: I learned more about Python and AR/Unity Jahlil: I learned about machine learning, how to use UiPath for automation, and how to create a web scraper Patrick: I learned how to create a node.js file, a JSON file, how to convert a CSV to JSON through UiPath, and basic automation through UiPath Vernicia: I learned more about machine learning, handling large data sets meaningfully and R
What's next for Disaster Emergency Relief
- Build a more comprehensive machine learning algorithm that takes more factors into account when determining if emergency aid will be necessary
- Link real time data through Firebase to allow immediate machine learning/recognition of potential disasters, and preemptively send out relief
Log in or sign up for Devpost to join the conversation.