A framework for multiple heterogeneous robots exploration using laser data and MARG sensor

We deal in this paper with the exploration problem using a team of heterogeneous robots. Since no assumption is made on the type of robot used, our system has to handle correctly in one single framework the specificities arising from each type of commonly used robots such as UAVs, ground vehicles or humanoids. To do so, we introduce a framework that handles efficiently the autonomous exploration problem for single robots then that allows sharing and fusing maps received from other teammates. Both theoretical background and software architecture are presented. We show experimental results taken for the case of a humanoid robot and a simulated UAV robot cooperating in previously unknown environment.