Nowadays the road transportation system is enlarged much due to the population and fast-moving world. Due to human fault, vehicle accidents are quite unavoidable. The primary cause for accidents is the desecration of traffic rules such as over speed, unplanned overtake, etc. To increase the safety of vehicle rider the autonomous cars are coming into the picture. The autonomous vehicles can decide on their own for safe self-driving. But, the autonomous vehicles are still on the bench because lots of realistic considerations and predictions are needed to make the system more intelligence. The vehicular adhoc network acts an important role in autonomous vehicle building by considering various facts. The motive of the research is to give assistance message to the drivers for safe overtaking. In the literature, safety assistance is provided through the Road Side Control Units (RSCU). But the construction cost and maintenance for the RSCU is too high. Our methodology will give a sustainable solution for safety assistance when the vehicle tries to overtake. The proposed methodology uses Cooperative Overtaking Sight Assistance (COSA) which predicts the comfort zone and danger zone boundaries for safe overtaking. The methodology greatly reduces infrastructure dependency. The experimental evaluations are carried out in SUMO and NS2.