Autonomous car, as the name suggests, a car that can drive on its own is the very concept that is being discussed in the major machine learning concepts. The basic approach for these types of projects is based upon the machine learning methods. Now is the time for the big advancement in this sector and is being in the process too in many of the developed and the developing countries. The goal of our project was to basically train a deep network to understand the steering behavior while driving, to make a fully autonomous car on the stimulator but being on a very realistic approach. For the purpose of the project, the deep network we are using is basically taking data from the front camera being mounted on the top of the car which basically predicts the steering direction at every moment. our system automatically learns about the traffic features and the traffic signals while driving for the almost perfect recognition of the markings and full autonomous driving.