🤖 AI Summary
Behavior trees (BTs) suffer from poor scalability and insufficient domain knowledge in robotic task planning, while large language models (LLMs), though powerful in reasoning, lack reliability and safety guarantees in planning. This paper proposes the first LLM-driven heuristic BT expansion framework: it leverages LLMs to generate semantically plausible heuristic paths that guide efficient BT construction; introduces two variants—optimal and satisficing heuristics—and incorporates action-space pruning and iterative reflective feedback to enhance accuracy and robustness. We theoretically prove that the framework ensures bounded planning complexity. Experiments across four service-robot datasets demonstrate significant improvements over baseline methods: faster planning, higher success rates, and guaranteed reliability and real-time performance.
📝 Abstract
Behavior Trees (BTs) are increasingly becoming a popular control structure in robotics due to their modularity, reactivity, and robustness. In terms of BT generation methods, BT planning shows promise for generating reliable BTs. However, the scalability of BT planning is often constrained by prolonged planning times in complex scenarios, largely due to a lack of domain knowledge. In contrast, pre-trained Large Language Models (LLMs) have demonstrated task reasoning capabilities across various domains, though the correctness and safety of their planning remain uncertain. This paper proposes integrating BT planning with LLM reasoning, introducing Heuristic Behavior Tree Planning (HBTP)-a reliable and efficient framework for BT generation. The key idea in HBTP is to leverage LLMs for task-specific reasoning to generate a heuristic path, which BT planning can then follow to expand efficiently. We first introduce the heuristic BT expansion process, along with two heuristic variants designed for optimal planning and satisficing planning, respectively. Then, we propose methods to address the inaccuracies of LLM reasoning, including action space pruning and reflective feedback, to further enhance both reasoning accuracy and planning efficiency. Experiments demonstrate the theoretical bounds of HBTP, and results from four datasets confirm its practical effectiveness in everyday service robot applications.