We wanted to create an app that would have an impact on the world. We were also surprised at the lack of third-party app integration in health-related apps, so the perfect solution was a marriage between the two.

The app is operated using Morse Code. Signals are translated into letters, which form strings of words. Certain words (POST, WATCH, TWEET, AMAZON, etc.) alert the program to in-app purchases and other actions, allowing users to search words, make purchases, watch videos, and post to social media.

We used Azure to deploy the initial app, natively built into the web application in Javascript. We then converted the files to Swift before creating a JSON through ManifoldJS.

We initially tried to build an OpenBCI-based application that would use the P300 library to turn signals from the brain into letters. Because of the number of bluetooth devices in the arena, we were forced to abandon the project, which would have only worked in the absence of all network signals (the noise would not allow for a clean reading necessary to calibrate the device for our P300 Speller). Thus, we chose to deal with muscle movement as opposed to simply brain waves.

We were able to reverse-engineer the Morsel API to work for a web application. We then imported code we wrote, as well as reverse-engineered APIs from Facebook, Twitter, etc. to simplify the functionality of the app.

We learned a lot about polyglot applications, as well as the difficulty of building chemical hardware in a 48 hour hackathon (especially if so many hardware hacks are taking place simultaneously.) Eventually, after converting our native web app to Swift, we used ManifoldJS to additionally create a JSON that would allow us to export our application to Windows and Android devices as well.

We're hoping to continue developing the features, and hopefully find a P300 library for Neuromore (it has yet to be invented) to continue our research in machine learning brain waves to form words.

Built With

Share this project: