Fruble

Safety is never guaranteed, with Fruble, users will have another form of safety on them at all times. This application gives users a discrete way to alert select contacts if they are in distress.

Comment

Inspiration

Our inspiration comes from a desire for better personal safety. If someone finds themselves in an unfortunate situation where they cannot use their phone due to immediate danger. We want to be able to contact authorities or close friends/relatives without having to call on the phone and be too obvious.

What it does

After launching the application, a user will have a preset safe phrase. While the application is open, it listens to audio through the microphone. While this is happening, it translates the input audio to text and tries to find the safe phrase of the user's choosing. If it is found, the application sends out emails to contacts of the user's choosing (setup upon installation). These emails include preset information detailing the fact that the user is in distress and needs assistance. The design allows for minimal contact with the device.

How we built it

We built Fruble using Android Studio for our frontend and a customized Python API using the SpeechRecognition library for our backend. With the assistance of an Android emulator, we were able to dial in the UI to a degree of our liking.

Challenges we ran into

Due to the fact that this was (all) our first hackathon, we have never coded an application this intense before. Nobody in our group had any experience with Android Studio and Speech Recognition. Setting up our environment took about eight hours, which left us feeling a little perturbed, but in the end, it was very satisfying to overcome the challenges. We had to spend lots of our time researching our way around Android Studio. Everyone has faced a seemingly immovable wall that, when you understand how to get around it, seems so unbelievably simple.

Accomplishments that we're proud of

After staying up all night (It was interesting) we had learned more in the 24 hours at this hackathon than all four of us combined in a month at college. James did an extremely good job at creating the UI and figuring out how to incorporate emailing (without Twilio which would have made it a breeze). Chris is proud of the functionality that we were able to produce as first timer, including adding the margin of error and scattered detection problems. Jackson overcame many challenges with the functionality of the backend of the UI in Android Studio, along with facing Gradle for the first time and figuring out connecting the front and backend. I (Ren) tangled with the speech recognition library, and I found myself feeling very accomplished when it finally correctly detected phrases that I was giving it.

What we learned

As mentioned in the previous questions, along with some new things, we learned new technologies and libraries along with bigger lessons. Those lessons include are as follows. It's probably good to sleep at at hackathons versus pulling all nighters, consuming more caffeine, does not in fact, keep you good throughout the night, and this will definitely not be our last hackathon. (Fruble returns)

What's next for Fruble

We have plans to develop this application so that it can continuously listen in the background (at user discretion) to increase the discreteness in terms of emergency help detection. We also want to incorporate more efficient language models so that our rate of accuracy is better for reading in audio. This will be activated with a "wake up" phrase like Apple Siri or Amazon Alexa. Deploying it to the app store is also on our radar and we have intentions to make it available for IOS depending on public and market responses. Getting in contact with law enforcement is also a potential path. This, along with incorporating location tracking, would allow law enforcement to quickly and effectively find someone who has gone missing or needs assistance.

Built With

Share this project:

Updates