case study 1

driver drowsiness detection

the problem

The last quarter of 2017, the leadership at Ola decided to explore solutions that ensures rider/driver safety on the roads. When we looked at data related to drivers' behaviour, we found that most of our drivers continuously took rides for more than 18 hours a day or more. I wanted to know if this behaviour makes our drivers drowsy and as a consequence compromise safety of themselves and others.


When I looked at research papers and accident data for India, I found out that it didn't exist. Then I explored data about other countries and came across a treasure trove of information to my surprise. The National Highway Traffic Safety Administration in the United States had extensive data on drowsiness related accidents.

Losses due to drowsy driving.

Human circadian rhythm.

A study comparing drowsiness with blood alcohol concentration.

We found out that drowsy driving is a real problem.

My Role

I was the sole designer and the developer on this project. I led the research, designing of a solution and developing the android application as a proof of concept.

The Solution

I decided to make an android application as a proof of concept as Ola driver devices are android based. I used facial recognition technology to track visual cues on a human face to detect drowsiness and sound an alarm if the driver is found to be drowsy.

Prototype in action