A metric approach for environments mapping

One of the main issues in mobile robotics is the autonomous navigation of a mobile robot in an unknown environment. Concurrent mapping and localisation or simultaneous localisation and mapping is a stochastic map building method which permits consistent robot navigation without requiring an a priori map. The governing idea which guides autonomous robotics consists in saying that the vehicle builds its chart progressively during exploration enabling it to evolve in the long term in unknown places in advance. When the robot environment chart is not known a priori, a generation module of incremental chart must obligatorily be integrated into the navigation system. The map is built incrementally as the robot observes the environment with its on-board sensors and, at the same time, is used to localise the robot. Unfortunately, the inaccuracy of the odometric sensors does not allow a sufficiently correct positioning of the robot. In this paper, simultaneous localisation and map building is performed with a metric approach which permits both precision and robustness. The most important innovation of the approach is the way how errors in the robot localisation control are handled by map building using the landmarks localisation information. The method uses data from a laser scanner to extract distances and orientations of landmarks and combines control localisation and metric paradigm. The metric approach, based on the Kalman filter, uses a new concept to avoid the problem of the drift in odometry. The simulation section will validate the maps representation approach and presents different aspect of environments.