Towards a decentralised mobile robot learning system for indoor environments

Learning in robotics has become a requirement in some advanced industrial applications since robots are expected to perform complex, and in some cases multiple, tasks while simultaneously handling the unpredicted conditions of the environment. The research discussed in this paper focuses on the development of an algorithm that will allow a mobile robot to determine its own physical dimensions. This decentralised control approach is necessary in flexible and reconfigurable industrial environments due to their dynamic realities. The paper also discusses the use of a popular robotic middleware, namely ROS, together with the Stage simulator package for algorithmic simulations and analysis.