Pandemic Defense

Inspiration

We've all experienced the oncoming feelings of fear and dread accompanied by a new variant coming, and when we just heard the news about Delhi being infected with a sub-variant of Omicron, we wondered how long it would take before the variant came here, to the US. Our lives had been halted due to our rising concerns, and this gave birth to Pandemic Defense, which allows us to prepare for a pandemic in various ways and protect our friends and family.

What it does

This app works in 4 parts: a map, a mask recommender, a CDC social guidelines informer, and a chatbot. The map allows you to look at which area you are from, wherever you are, and see the risk you have due to an oncoming variant, which has been determined due to an AI/ML Algorithm based on ICU Hospitalizations. The mask recommender recommends specific kinds of masks depending on the severity of a pandemic in your location, varying from cloth masks to the N95 Series. The social distancing recommendations come directly from the CDC, recommending certain practices based on the severity of a pandemic.

How we built it

We built this project using XCode, Python, SKLearn, PythonKit, and CSVs. We used PythonKit to tie our backend and frontend together, while XCode and Swift served as our UI. We used CSV files for our databases and we used SKLearn for our Regression Model.

Challenges we ran into

A challenge we ran into was tying together our chatbot (generated in Python) to our frontend (generated in Swift). We spent hours of research until we found PythonKit, and since we had to transfer data and variables in large amounts, we had to optimize extremely. We're proud of the fact that we overcame this challenge with PythonKit and genius uses of Structs and other data structures.

Accomplishments that we're proud of

We're proud of our MapKit UI, as we have implemented a lot of advanced systems there. We used complex gradients of colors that are custom made, used a database from Kaggle to get a list of countries, and used Artificial Intelligence to generate risk indices for countries. We also used regression models for predictions of cases rising and dropping, and our analyses show that our data is extremely accurate.

What we learned

We learned a lot about communication and collaboration, as we realized that some of us are better at some parts of this project than others, and proper delegation is key to success.

What's next for Pandemic Defense

For Pandemic Defense, we plan on implementing a Neural Network for our Chatbot so it acts more humanely, as well as adding means to identify more diseases than just COVID. We want our Pandemic Defense App to apply for more than just COVID, and we also wish for our app to be connected to a live database through AWS so that it can work in the future.