Behaviour Tree-Based Task Execution Controller for Internal Transportation Robots

The aim of this work is to present the structure and operating principle of a high-level controller for receiving, interpreting and executing tasks for autonomous mobile robots. The controller is based on behaviour trees. The py_trees library was used to develop it. The algorithm has been applied to the mobile robots designed for internal transportation. Special emphasis was put on the description of the incoming tasks handling procedures. It is explained how they were designed and how the behaviours are interpreted by the robots. The novelty of the proposed solution is the static structure of the controller, which, compared to dynamic approaches, does not require reconfiguration of the tree to handle different types of tasks. In the paper, the authors compare their own solution to methods previously used in mobile robotics. In conclusion, the advantages of the concept presented in the paper are listed.