Autonomous robots are extensively appearing in agricultural domain with ranges of application, and still evolving. Navigation is one of the most important aspect in designing an autonomous robot. Therefore, the present study focuses the navigational strategy for an orchard robot using model based design. The navigational model is developed in MALTLAB/Simulink as a software development and simulation platform. This navigational unit includes motion planning, obstacle avoidance and data transmission module. The navigational model acquires planned path coordinates from path planning module required in motion planning algorithm. The pure pursuit controller is used in this study for the path tracking algorithm. Before hardware testing, the model was verified in MATLAB 2D virtual environment to visualize the robot navigation. The geo-referenced image was used for localization. This increases the reliability for navigation in GPS denied environment. The navigational control strategy for a real robot is verified using the remote controlled robotic platform named elWObot. The developed model was tested in the laboratory with operating speed of 1 m/s and was able to navigate successfully. Additionally tests were carried out to validate the model performance in outdoor environment.