Beyond Predefined Actions: Integrating Behavior Trees and Dynamic Movement Primitives for Robot Learning from Demonstration

2025-05-13

AI Summary

Existing robot skill transfer methods suffer from inherent limitations: behavior trees (BTs) rely on expert-designed low-level actions, while dynamic movement primitives (DMPs) lack high-level task logic. Method: We propose the first end-to-end joint learning framework that embeds DMP controllers within a BT architecture, simultaneously optimizing both BT structure and DMP parameters—without pre-defined actions or human priors. Leveraging inverse reinforcement learning and hierarchical policy learning, our approach acquires interpretable, modular, and editable skill representations from a single human demonstration. Results: On a multi-task robotic manipulation benchmark, our method achieves 92% demonstration reproduction accuracy and reduces policy editing time by 76%. It enables cross-task component reuse and real-time online adaptation, effectively overcoming the respective bottlenecks of BTs and DMPs.

Technology Category

• Intelligent Robots: Behavior Learning & Control
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Interpretable policy representations like Behavior Trees (BTs) and Dynamic Motion Primitives (DMPs) enable robot skill transfer from human demonstrations, but each faces limitations: BTs require expert-crafted low-level actions, while DMPs lack high-level task logic. We address these limitations by integrating DMP controllers into a BT framework, jointly learning the BT structure and DMP actions from single demonstrations, thereby removing the need for predefined actions. Additionally, by combining BT decision logic with DMP motion generation, our method enhances policy interpretability, modularity, and adaptability for autonomous systems. Our approach readily affords both learning to replicate low-level motions and combining partial demonstrations into a coherent and easy-to-modify overall policy.

Problem

• Integrate BTs and DMPs to overcome their individual limitations
• Learn BT structure and DMP actions from single demonstrations
• Enhance policy interpretability, modularity, and adaptability

Innovation

• Integrates Behavior Trees with Dynamic Movement Primitives
• Learns BT structure and DMP actions jointly
• Enhances policy interpretability and modularity