ResQ | Devpost

Inspiration

What it does

ResQ is a holistic natural disaster recovery tool. -Officials (FEMA, Red Cross, etc.) can manage their resource centers on a WebApp: by managing the distribution of resources, the location of distressed individuals, and publicly sourced alerts.
-Private Citizens can use a mobile app as a one-stop-shop solution for their emergency needs including: finding the nearest medical service, food/water source, or shelter source, alerting officials of power outages, fallen trees, or obstructed roads, and finally trigger and emergency response. Users can use augmented reality to point them towards the closest resource. -Emergency Response Teams: can be dispatched through the app to find private citizens. A convolutional neural network processes Arial imagery to tag and find distressed citizens without cell service

How I built it

The WebApp is built in React.JS. The mobile app is built with swift in Xcode. The backend was made using Firebase. The AI/Deep Learning portion was built using Keras.

Challenges I ran into

We ran into several challenges throughout the course of this project. We all dealt with some ideas and technologies we had very little experience with before. Implementing the AR/VR was challenging as this technology is very new and is still very hard to use. Using a pretrained neural network to do image detection (drawing the boudning box) was very challenging as it is a machine learning problem we had never tackled before and one in which very little documentation exists. Also, dealing with many of the sponsor APIs was initially very challenging to deal with as some of the endpoints were hard to successfully interact with.

Accomplishments that I'm proud of

We think that the scale of this project is huge and has a tremendous amount of application in the real world. This app (on the mobile side) gives people who are victims of a natural disaster a place to report their location, find any resources they may need, and report anything potentially dangerous. This app also (on the web side) gives rescuers a central database to locate and keep track of people who are currently in danger. Lastly, this app also uses deep learning to use drones to identify stranded humans who may not have cell service. We are truly proud of the scale this project achieves and all the rich and various technologies involved.

What I learned

We all learned various skills in our respective parts of the application: React, iOS with AR/VR, Firebase, Keras.

What's next for ResQ

The next steps would be to actually implement the deep learning portion of the project, preferably with a drone that could transmit a constant stream that could be processed to see if there are any humans in a certain area and transmit their coordinates appropriately. We also want to build out each specific feature of the mobile app, including directions to water, food, shelter, hospital, or gas.