The progress in technology has made vehicles safer and the quest to make them even more safe is never ending. Autonomous cars present the solution to make cars much more safer by eliminating the primary cause of road accidents, human error. However, autonomous cars tend to fail in decision making especially in complex traffic environment prompting human intervention, since the technology is not yet mature. The transition to autonomous cars can be achieved via teleoperated driving which allows human operator to remotely control the vehicle via mobile network. This thesis presents Model Predictive Control (MPC) based driver assistance system for semi-autonomous teleoperated road vehicle that helps avoid collision with static and dynamic obstacles. This system aims to mitigate some of the key challenges in teleoperated driving like reduced situational awareness and latency. The proposed system posses the ability to correct lateral and longitudinal motion of the vehicle and explicitly defines its authority to override human operator. To enhance trust in the human operator over the system, visual feedback of the vehicle behaviour is proposed as an additional set of information to the human operator. Simulations were done using high fidelity vehicle model and the results validate the expected behaviour of the system, as designed for teleoperated driving system setup of Institute of Automotive Technology, Technical University of Munich. The designed system is ready for implementation in the actual experimental vehicle and hence real tests can be conducted.