A framework for simulations and tests of mobile robotics tasks

Abstract

This paper presents an education framework, developed in Matlab, for studying and experimenting typical mobile robotics tasks such as obstacle avoidance, localization, navigation and SLAM. The most important characteristic of this framework is the ability to easily switch from a simulator to a real robot to tune and test algorithms and to evaluate results in simulated and real environments. The framework is being used with interesting results in robotic courses at the Universita Politecnica delle Marche in Ancona, Italy. In the second part of the paper a test case to evaluate an optimization of a Monte Carlo Localization process with sonar sensors is presented.