An autonomous vehicle or a self-guided vehicle (SGV) is a vehicle (robot) that performs the desired tasks in an unstructured environment without continuous human guidance. Almost all applications of an SGV require a vehicle that is capable of moving accurately and repeatedly to a particular location within its environment while executing a specific task. The accuracy and robustness in performing a specific task are therefore very important for the SGV to achieve a high level of performance. This paper introduces a new spherical velocity motion model and a new spherical odometry-inertia motion model for 3-D local landmark-based autonomous navigation. These new models are high accuracy and low-cost models. As modeling the contents of the immediate environment is fundamental, estimation of the position of the vehicle with respect to the external world is fundamental as well. Hence, using the most powerful tools of estimation theory, i.e., the extended Kalman filter (EKF) and the unscented Kalman filter (UKF), which give the best estimations in noisy environments, will prove the accuracy and robustness of these 3-D models.