The precise control of mobile robots is not an easy task. Because of many reasons, we can experience that the path we wanted the robot to go through is different from the path which is followed by the robot in the real life. Despite we can create a mathemati-cally perfect description using the wanted speeds of the wheels and the time, it is not enough. These errors can be caused by the sub-optimal behaviour of the motors, or the interaction between the robot and the environment. Naturally we can find many other reasons, but these are the most important ones.
I would like to solve these problems with the modern inertial sensor technology, by tracking the actual movement of the robot, and from the sensor data calculating and trying to correct the errors. In this thesis I am working specifically with a gyroscope and trying to compensate the errors in the orientation of the robot.
At the first part, I would like to offer a small insight into the world if gyroscopes based on external resources. The main goal of this part is to provide some important or just interesting things about these sensors and their application. Then I show how I imple-mented a real-time determination of the actual orientation of the robot based on data from a gyroscope, particularly with regard to the data processing and the solutions of the problems. In the next section I present how to use this information in a robot with differential drive to develop a controller, which is able to detect these errors and make corrections. Finally, I present a brief description about the usage of the API which con-tains these functions.