A new method for a driver in the loop vehicle control is proposed in this paper. The driver is modeled as a linear optimal controller with time varying delay, who uses road preview information as well as dynamic behavior of the vehicle to steer the car and follow the desired path. The delay block represents the cumulative drivers observation and the action delay. Since regular controllers do not have access to the future road preview data, the road preview information is modeled as unknown bounded uncertainty. A robust state-feedback controller based on Lyapunov-Krasovskii methodology is designed to mitigate effects of delays and uncertainties. The controller in this scheme only uses the current states of the vehicle, driver input, and the driver model. Simulation results demonstrate the effectiveness of the proposed method.