There are enough tools for resource distribution at sites of natural disasters, but I have yet to come across a project that considers the human lives at stake when networks crumble and loved ones go missing.
What it does and how I built it
PHIND (pronounced "find") aims to create a mesh network that allows people in zones heavily affected by natural disaster to communicate without traditional network support. Along with automatically translated local chat, computed by Google Cloud Platform, we also use machine learning and computer vision with Google Cloud's AutoML to look at images of lost persons and compare them with images of persons reported to be found in order to rejoin families and friends. To expand past the local mesh network, and option is available to share images of lost loved ones or found survivors to snapchat through SnapKit's Creative Kit. The mesh network is available on both mobile and the web to make the platform as accessible as possible.
Challenges I ran into
One of the largest challenges I faced in building this project was the lack of time — I was at an algorithm's competition till around 6pm and didn't start hacking till 11pm, burning half of the hack time available. Beyond that, my team had to leave at 12am and did not return till late morning.
I also had no experience with swift or iOS development prior to HackUNT, so getting a basic app up and running was challenging enough. After implementing SnapKit and hooking into different APIs such as GCP, I feel like I got the hang of iOS and will definitely try to sharpen my skills at a future hackathon.
Accomplishments that I'm proud of
1 man team at a 4 person hackathon. 12 hours at a 24 hour hackathon. And I ended up with a project that is not just viable but unique, useful, and fun to build.
What I learned
I learned how to develop iOS apps, work with external libraries with Cocoapods, use SnapKit to integrate with Snapchat, and what a mesh network is.
What's next for Persistent Household Indicator for Natural Disasters (PHIND)
Only time will tell, and right now I have time to sleep. Bye bye.