Combining Topological and Metric: A Natural Integration for Simultaneous Localization and Map BuildingAbstract

In this paper the metric and topological paradigm are inte-grated in a single system for both localization and map build-ing. A global topological map connects local metric maps, allowing a compact environment model, which does not re-quire global metric consistency and permits both precision and robustness. Furthermore, the approach permits to han-dle loops in the environment by automatic mapping using the information of the multimodal topological localization. The system uses a 360 degree laser scanner to extract corners and openings for the topological approach and lines for the metric method. This hybrid approach has been tested in a 50 x 25 m2 portion of the institute building with the fully auton-omous robot Donald Duck. Experiments are of three types: Maps created by a complete exploration of the environment are compared to estimate their quality; Test missions are randomly generated in order to evaluate the efficiency of the localization approach; The third type of experiments shows the practicability of the approach for closing the loop. 1