To try it out, download the apk below:
Note: your device must have a large amount of RAM.
https://drive.google.com/open?id=1im6fOyNT9mNpDHgg_wplihe8bzB9hk78
Screenshots
Inspiration
We wanted to tackle a real problem that affects many people everyday. We wanted to create something to help keep everyone safe and save lives. We also had an interest in ML and APIs and wanted to built something with those. We found that many people are affected everyday by fatigued driving. In fact, it is estimated that 26% of accidents involve fatigued driving. So, we decided to try to help reduce this number by building our app, HootGuard.
What it does
HootGuard is meant to watch drivers using the front facing camera (as many drivers place their phones on the dashboard of the car anyways) and alert them when they close their eyes for prolonged periods of time, indicating that they may have fell asleep.
How we built it
Our app uses the CameraX library to take the image from the front facing camera. Then, we used the ML Kit in the Google Firebase API to look at the face of the driver and detect if they are dozing off. The app then plays a notification to wake up the driver and suggests that they take a break to rest.
Challenges we ran into
We faced many challenges, particularly bugs that we could not solve for hours as well as limitations that we discovered along the way. We worked around these limitations and worked together to debug as best we could and ended up with a working first version of our app.
Accomplishments that we're proud of
We built our first Android app and did our first integration of an API (Firebase API) with the ML Kit and ended up with a cool working first version.
What we learned
We learned a lot from this hackathon. All of us had never built an Android app before nor integrating APIs into programs, so it was a very good learning experience for us as we taught ourselves how to do this.
What's next for HootGuard
In a 36 hour hackathon, its difficult to have a fully ready for release app, so our goal for the app is to improve it by making it more efficient and responsive, as well as create an even better UI and implement more driving safety features, such as distraction detection.