MRBTP: Efficient Multi-Robot Behavior Tree Planning and Collaboration
Yishuai Cai, Xinglin Chen, Zhongxuan Cai, Yunxin Mao, Minglong Li, Wenjing Yang, Ji Wang
TL;DR
MRBTP tackles multi-robot BT planning by introducing cross-tree expansion to coordinate heterogeneous actions across BTs and intention sharing to enable parallel execution while maintaining robustness. The framework is proven sound and complete, with a polynomial worst-case complexity $O\left(|\bigcup_{i=1}^n \mathcal{A}_i||\mathcal{S}||\mathcal{L}|\right)$, and it is augmented by an optional subtree pre-planning plugin that leverages Large Language Models (LLMs) to generate long-horizon subtrees for faster planning. Empirical results in warehouse management and home-service tasks demonstrate MRBTP's robustness to action failure and its superior execution efficiency, especially when combined with intention sharing and LLM-generated subtrees; different LLMs show consistent gains with increased reasoning ability and feedback loops. The work also highlights practical benefits in reducing inter-robot communication while preserving task reliability, making MRBTP a scalable approach for real-world multi-robot coordination with interpretable BTs. Overall, MRBTP advances reliable, scalable multi-robot planning by integrating principled BT theory with cross-tree coordination, intention sharing, and AI-assisted long-horizon reasoning.
Abstract
Multi-robot task planning and collaboration are critical challenges in robotics. While Behavior Trees (BTs) have been established as a popular control architecture and are plannable for a single robot, the development of effective multi-robot BT planning algorithms remains challenging due to the complexity of coordinating diverse action spaces. We propose the Multi-Robot Behavior Tree Planning (MRBTP) algorithm, with theoretical guarantees of both soundness and completeness. MRBTP features cross-tree expansion to coordinate heterogeneous actions across different BTs to achieve the team's goal. For homogeneous actions, we retain backup structures among BTs to ensure robustness and prevent redundant execution through intention sharing. While MRBTP is capable of generating BTs for both homogeneous and heterogeneous robot teams, its efficiency can be further improved. We then propose an optional plugin for MRBTP when Large Language Models (LLMs) are available to reason goal-related actions for each robot. These relevant actions can be pre-planned to form long-horizon subtrees, significantly enhancing the planning speed and collaboration efficiency of MRBTP. We evaluate our algorithm in warehouse management and everyday service scenarios. Results demonstrate MRBTP's robustness and execution efficiency under varying settings, as well as the ability of the pre-trained LLM to generate effective task-specific subtrees for MRBTP.