The field of robotics is becoming increasingly more important and consequently, students need better tools to gain knowledge and experience with them. The University of Skövde was interested in developing a learning tool focused on a virtual simulation of mobile robots. Despite the fact that there are several programmes to create this tool, MATLAB was preferable because of its strong presence in educational institutions. The objectives were oriented towards testing different robot kinematics in an adjustable virtual 3D environment. Moreover, the simulation needed a part in which future users could design own algorithms in order to control the AGVs. Therefore, sensors such as LIDAR sensors were necessary to enable a possible interaction between the robot and the scenario created.
This project was developed with a previous study and a comparison of some MATLAB projects and tools. After that, the scenario and the simulation were produced. As a result, a virtual simulation has been created emphasising that the user could modify and adapt multiple parameters such as the size of the AGV, the form of the virtual environment or the selection of forward or inverse kinematics in order to develop different types of algorithms. Other features can be adjusted manually such as the type or number of sensors as well as SLAM conditions. Finally, this thesis was conducted to give a basis about mobile robots and to be a first step for operating with real robots. The simulation also provides an easy to use interface in which students can keep working in it through the introduction of new applications related to image processing or more sophisticated algorithms and controllers.