Framework using ROS and SimTwo simulator for realistic test of mobile robot controllers

Abstract

In robotics, a reliable simulation tool is an important design and test resource because the performance of algorithms is evaluated before being implemented in real mobile robots. The virtual environment makes it possible to conduct extensive experiments in controlled scenarios, without the dependence of a physical platform, in a faster and inexpensive way. Although, simulators should be able to represent all the relevant characteristics that are present in the real environment, like dynamic (shape, mass, surface friction, etc.), impact simulation, realistic noise, among other factors, in order to guarantee the accuracy and reliability of the results. This paper presents a ROS (Robot Operating System) framework for the SimTwo simulator. ROS is an open-source library that is commonly used for the development of robotic applications since it provides standard services and promotes large-scale integrative robotic research. SimTwo is a realistic simulation software suitable for test and design of several types of robots. This simulator conducts realistic navigation procedures, since the driving systems, the sensors, the mechanical and physical properties of the bodies are precisely odeled. The framework presented in this research provides the integration of ROS-based systems with the SimTwo simulator. Therefore, this framework reduces the risk of damage of expensive robotic platforms and it can be used for the development of new mobile robot controllers, as well as for educational purposes. © 2015 Springer International Publishing.