Self-Driving-Car

Inspiration

I wanted to explore how self driving cars worked. As it may be immediately obvious, I did not have a self driving car with me that I could inspect. So, I decided to make one myself - in a simulation ;)

What it does

My self driving car can detect lanes and predict the best steering angle at any given point depending on the path ahead.

How I built it

I used machine learning libraries like Tensorflow and Keras to build the machine learning part of the project (detecting lanes, predicting steering angle etc.) and socketio to connect the machine learning model with the Udacity Self Driving Car simulator to make the car full functional and able to drive itself.

Accomplishments that I'm proud of

This was a project that was well out of my reach and I was not even sure if I was ever going to finish it. However, some perseverance and help from the internet I was able to complete the project and publish it.

What's next?

I would love to extend this project in perhaps some other simulator where there are traffic lights, multiple cars, pedestrians and all the other complexities real world road traffic would have to make a full functional and able self driving car!

Built With

• jupyter-notebook
• python