In this research paper, our approach is to present a practical solution for all the road car accidents which takes place. Our model will monitor the preceding vehicles and simultaneously it will calculate the distance between the vehicles. Now, if the preceding vehicle gets too close, then our model willwarnthat driver of an impending crash. The driver of the vehicle will simply be alerted by any audible signal after reaching the threshold distance between the vehicles. This model will be very much helpful, especially for the learners. This calculation of the distance between the vehicles will be done by our CNN based model and a camera which will be located on the windshield to click the rear side images of the preceding vehicle in real time. In our experimental results, we can see that our model can calculate the distance and we would be able to achieve an overall good accuracy. In the past, researchers used these technologies like ADAS and FCW system which are entirely dependent on the hardware and cost-effective as well. So, we are using CNN in FCW system instead of using the hardware like radars. In this paper, the concept of Advanced Driver Assistance System (ADAS) and Forward Collision Warning (FCW) are also described.