The objective of this project is to encourage the way toward driving a vehicle via automating it. The aftereffect of this project will diminish the number of fender benders happening nowadays. The thought is to make an autonomous vehicle(a mixed-race model of Level two and three) that utilizes just some sensors(collision indicators, temperature locators and so on) and camera module to go between objections with negligible/no human mediation. The vehicle contains a prepared Convolutional Neural Network (CNN) which will predict all the boundaries that are needed for efficiently driving a vehicle. They are legitimately associated with the fundamental guiding component, and the yield of the deep learning model will decide the directing point of the vehicle.