Robotics has been introduced in education at all levels during the last years. In particular, the application of mobile robots for teaching automatic control is becoming more popular in engineering because of the attractive experiments that can be performed. This paper presents the design, development, and implementation of an algorithm to control the position of a wheeled mobile robot using Reinforcement Learning in an advanced 3D simulation environment. In this approach, the learning process occurs when the agent makes some actions in the environment to get some rewards. Trying to make a balance between the new information of the environment and the current knowledge about it. In this way, the algorithm is divided into two phases: 1) the learning stage, and 2) the operational stage. In the first stage, the robot learns how to reach a known destination point from its current position. To do it, it uses the information of the environment and the rewards, to build a learning matrix that is used later during the operational stage. The main advantage of this algorithm concerning traditional control algorithms is that the learning process is carried out automatically with a recursive procedure and the result is a controller that can make the specific task, without the need for a dynamic model. Its main drawback is that the learning stage can take a long time to finish and it depends on the hardware resources of the computer used during the learning process.