wakeTFup

Inspiration

With tired drivers being a constant threat to the safety of all of us, we would like to help take matters into our own hands and assist with public safety.

What it does

wakeTFup uses a Google-trained machine learning model to detect when drivers are tired by recording the amount of time they spend with their eyes closed. If they spend too much time with their eyes closed (longer than "just a long blink"), it will attempt to wake them up, and even dial an emergency contact if necessary.

How we built it

We used a combination of the Android Vision APIs, a Python server, a node.js server, and a bluetooth communication.

Challenges we ran into

Getting the bluetooth stack properly working on the Dragonboard was definitely difficult, but I'm glad we were able to get it taken care of.

Accomplishments that we're proud of

Getting everything working together was definitely our most proud accomplishment of this project.

What we learned

We learned how to properly utilize and interface with the Snapdragon 410c board.

What's next for wakeTFup

One future goal would be to get Android Auto integration fully working, if possible.

Built With

• 3d-printing
• android
• android-vision-api
• bluetooth
• debian
• domain.com
• google-cloud
• machine-learning
• node.js
• python
• qualcomm-dragonboard-410c