Multisensor data fusion for obstacle detection in automated factory logistics

This paper describes data fusion methodologies for obstacle detection in an automation system based on advanced Automatic Guided Vehicles (AGV), used for automated logistics in modern factories. We present the background of the problem, introducing generic aspects of the system architecture designed to cope with the obstacle detection in automated factory logistics; then, we focus on the system specification for the module responsible of integrating data from different sources and providing a global representation of the environment. Finally, we present a comparative analysis among different strategies of multisensor data fusion compliant with the requirements of the described system, highlighting their advantages and drawbacks.