Towards goal-driven behaviour control of multi-robot systems

Abstract

The task-level control of mobile robots in uncertain environments demand high adaptivity and cognitive capabilities that allow for flexible decision making and planning. This work introduces the ROS Hybrid Behaviour Planner (RHBP). RHBP combines the advantages of a reactive and adaptive behaviour-based layer with a formal symbolic planner that enables appropriate goal pursuance. Here, we present the first development stage of our system that enables more adaptive multi-robot decision making and planning in the widely used Robot Operating System (ROS) ecosystem.