This article studies the most common methods of autonomous vehicle guidance, as well as a review of companies that have already developed their own self-guided tractors. The most common methods for autonomous guidance are through the position obtained from an RTK-GPS and by artificial vision. For both cases, sensors have to be implemented that help the task assigned, such as LIDAR sensors, proximity sensors, among others, that detect the tractor to know the working environment and define the correct route and even avoid collision, to comply with the assigned task, managing to improve production time and quality. The two guidance methods mentioned in this work fail to have error = 0, either by the type of sensor or by the technique used. Among the most commonly used techniques for obstacle avoidance are the possible field method or the probabilistic route planner.