Self-driving cars have become more and more discussed in the recent years. Since major innovations always start with small steps, it is advisable to project the complex process of developing such a self-driving vehicle, which takes place at a very high and cost-intensive stage, onto a smaller and more manageable level. This paper will show the process of planning, building, and programming a self-driving robot that will be able to clear simple labyrinth-like structures and perfect mazes autonomously. At the beginning, a few fundamental components like the used microcontroller or the quadrature encoder are explained. In the following, experiments on the output of the used infrared sensor are conducted, the data is analyzed and the tendency to an error in the measurements is then calculated. Since the robot uses the drive mechanism known as differential drive, this principle will also be explained. Further, the building process of the robot is described. Furthermore additional tests are run on the program for determination of the position which can then be used to map the robot’s path while driving. The end of this research work will be about the implementation of different kinds of driving methods and algorithms which will allow the robot to maneuver through simple structures. Here, the difference between an open-loop controller and a closed one is discussed and evaluated.