Ultra-Wide Band (UWB) sensors are innovative devices constructed for efficient wireless communications that have recently being used for vehicle localization in indoor environments. In contrast, GPS sensors are well-known satellite-based positioning devices widely extended for outdoor applications. We evaluate in this paper the combination of both technologies for efficient positioning of vehicles in a mixed scenario (both indoor and outdoor situations), which is typical in applications such as automatic guided vehicles transporting and storing goods among warehouses. The framework we propose for combining sensor information is Monte Carlo Localization (also known as Particle Filters), which is a versatile solution to the fusion of different sensory data and exhibits a number of advantages with respect to other localization techniques. In the paper we describe our approach and evaluate it with several simulated experiments that have yielded promising results. This work, supported by the European project CRAFT-COOP-CT-2005-017668, becomes a first step toward a robust and reliable localization system for automated industrial vehicles.
展开▼
