A decentralised system approach for controlling AGVs with ROS

Abstract

The current industrial state of the art for automated guided vehicles relies on centralised controllers dispatching transports to vehicles along predefined paths. The design of these paths is time-consuming and has to be done in advance, followed by an extensive testing phase. In the field of mobile robotics, robust path planning and navigation algorithms exist. However, they have not yet found their way into industrial applications in larger numbers. In this paper, we present a system architecture for a decentralised control of multiple automated guided vehicles performing material transportation tasks in intra-logistic applications which is based on mobile robotics solutions. The proposed system includes solutions for self-localisation, behaviour control, conflict-free routing and motion control. The non-centralised control of the system architecture allows for dynamic path planning and traffic coordination. Its implementation is based on the Robot Operating System, the de-facto standard middleware for robotics applications. We give an overview of the overall system architecture as well as the coordination mechanisms and show a first proof of concept in simulations.