According to the National Sleep Foundation’s 2005, 60% of adult drivers say they have driven a vehicle while feeling drowsy in the past year, and more than one-third, have actually fallen asleep at the wheel. Drowsiness is one of the key causes of driving accidents. In 2013, drowsy driving was responsible for 72,000 crashes, 44,000 injuries, and 800 deaths according to The National Highway Traffic Safety Administration.

We wanted to develop a health and safety hack which prevents car accidents caused by drowsiness and benefits our society.

What it does?

DriveWake is a drowsiness prevention mechanism which detects the driver's drowsiness levels. It uses facial recognition features of Microsoft Kinect technology, and alert the driver with vibrations and sounds.

How we built it

There are three aspects to this advanced system. Microsoft Kinect handles the facial recognition by analyzing blinking patterns and relays an indication of drowsiness level to a web server. The server notifies an android application when the driver is drowsy. Then, the android app prompts the vibrations on Microsoft Band and plays sound alerts on the phone.

Challenges we ran into

The sensor in Microsoft Kinect is sensitive to small changes in eye movement. This was interfering with our results since it would even detect eyeball movement when the eyelids were closed. We also faced issues with client server interaction between the web server and the Android application.

Accomplishment that we're proud of

Learned how to use Microsoft Band, Azure, and Kinect for the first time ; improved our android development skills; survived with minimal sleep


Ensure safe driving by creating a warning mechanism for sleepy drivers. This project uses Microsoft Kinect to detect facial signs which indicate that a driver is falling asleep and gives a warning to the drowsy driver.

Share this project: