Maintaining Consistent Geographic Description of the Environment of an Autonomous Mobile Robot

Abstract

In this paper we present a method for a mobile robot to explore autonomously an unknown environment. many issues are involved by such a task. In particular we present a model to represent geographic knowledge, based on an extension of the “Diktiometric representations”; this kind of representations broaden the topological model to include the geometric relations between places. The model we propose is a network of 2D geometric descriptions connected by arcs with geometric relations between nodes. We paid special attention to the maintenance of this model, providing mechanism to allow the consistent fusion of sensory observations. To keep low uncertainty inside each node, we introduce in the fusion process “internal relations”, whose measurements are affected only with the uncertainty of the sensor system and not on the robot location.