Behavior Trees based Flexible Task Planner Built on ROS2 Framework

As modern industrial robotic systems become smarter and more flexible, they are rather tailored for specific, large-scale applications, making its implementation too complex and costly for smaller operators. The ACROBA project aims to develop and demonstrate a novel concept of cognitive robotic platforms based on a modular approach able to be smoothly adapted to virtually any industrial scenario applying agile manufacturing principles. The flexible task planner introduced here is developed as part of this European project. It aims at providing a user-friendly tool allowing the user to reconfigure the robotic cell in quick and cheap way. The main characteristics of the task planner are the behavior tree-based control (which avoids the implementation of complex Functional States Machines), the embedded HMI interface (providing basic functionalities like start, pause, resume, manual selection of skill to run…), the embedded PDDL solver (which allows the task planner to automatically rewrite the task at run time based on factory or sensors inputs), an easy interface with a GUI to design the task manually, and a work-in-progress Assembly Sequence Planner (ASP).