Software to Alert Distracted Drivers

Ojaswy Akella (~Ojaswy)


What is Distracted Driving? Distracted Driving is when an activity of any sorts distracts a person from his primary duty of driving. Activities include the usage of Mobile Phone, eating and/or drinking, watching videos, conversation, adjusting the music player, and but not limited to even using the GPS while driving.

There are three types included:

  • Visual Distraction - Taking your eyes off the road
  • Manual Distraction - Taking your hands off the wheel
  • Cognitive Distraction - Taking your mind off driving

Why should this be addressed? Distracted Driving has been identified as a major threat on the road. Statistics show that usage of Mobile Phones while driving is the major culprit in distracting drivers which causes their attention to be off the road and later cost their life.

India alone accounts for 10% of the Global Road Crash failures that is 146,133 people, out of which there were 12,589 children and these are the statistics as of 2015 and they have to change and that exactly is what this projects wishes to accomplish.

How does this work? I am developing a model which includes the usage of a camera to observe the driver at all times. From a dataset, I train the model to identify what the driver is doing and alert him irrespective of the surroundings. You can think of it as the beep sound which occurs when the seatbelt isn't put on. This beep would be synchronised with the braking system and effectively halt the vehicle if the beep persists.

This would be accompanied by a live demo which would depict the functionality.

The libraries I am using are Keras, Tensorflow, PILLOW, scikit-image, NumPy, OpenCV and testing out various models and their efficiency.

Draft Poster -


A vague idea of Image Processing would better help understand the functioning of the model.

Speaker Info:

Hello, I am Ojaswy Akella, I am 19 years old and currently majoring in Computer Science from University of Hyderabad. I am a major Machine Learning enthusiast and I aim to become a top notch researcher in the same. I started learning Python about 2 years ago and once I realised the power it holds, it has been a joyride since. I have previously worked on many Data Science and ML projects, my very first being a Chatbot based on Artificial Intelligence. I have interned at the Indian School of Business this summer on developing a prototype of a Financial Portfolio which delves into the application of AI in Finance. I have also worked on developing an Automated Driving System and had it deployed and tested on toy cars. You can find my other projects in the GitHub profile link I left below.

Speaker Links:

GitHub -

Portfolio -

Id: 1538
Section: Data Science, Machine Learning and AI
Type: Poster
Target Audience: Beginner
Last Updated: