One Demo Is All It Takes: Planning Domain Derivation with LLMs from A Single Demonstration

2025-05-23

AI Summary

Traditional Task and Motion Planning (TAMP) frameworks rely on manually engineered symbolic planning domains, incurring high annotation costs and poor generalization—severely limiting practical long-horizon task planning. This paper introduces the first fully automated method to derive end-to-end usable planning domains from a single annotated physical execution demonstration, requiring no predefined predicates, action templates, or domain knowledge. Our approach synergistically integrates large language model (LLM)-driven symbolic reasoning, physics-based simulation, trajectory annotation parsing, and a TAMP framework to achieve seamless coupling between high-level symbolic planning domains and low-level motion skills. Evaluated across nine diverse environments comprising over 1,200 tasks, our method significantly outperforms six LLM-based planning baselines, reducing token consumption by 37%. Furthermore, it has been successfully deployed on multiple real-world robotic platforms, demonstrating robustness and practical applicability.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Multiagent Systems: Multiagent Planning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information

Abstract

Pre-trained Large Language Models (LLMs) have shown promise in solving planning problems but often struggle to ensure plan correctness, especially for long-horizon tasks. Meanwhile, traditional robotic task and motion planning (TAMP) frameworks address these challenges more reliably by combining high-level symbolic search with low-level motion planning. At the core of TAMP is the planning domain, an abstract world representation defined through symbolic predicates and actions. However, creating these domains typically involves substantial manual effort and domain expertise, limiting generalizability. We introduce Planning Domain Derivation with LLMs (PDDLLM), a novel approach that combines simulated physical interaction with LLM reasoning to improve planning performance. The method reduces reliance on humans by inferring planning domains from a single annotated task-execution demonstration. Unlike prior domain-inference methods that rely on partially predefined or language descriptions of planning domains, PDDLLM constructs domains entirely from scratch and automatically integrates them with low-level motion planning skills, enabling fully automated long-horizon planning. PDDLLM is evaluated on over 1,200 diverse tasks spanning nine environments and benchmarked against six LLM-based planning baselines, demonstrating superior long-horizon planning performance, lower token costs, and successful deployment on multiple physical robot platforms.

Problem

• Automating planning domain creation from single demo
• Reducing manual effort in symbolic action design
• Enhancing long-horizon task correctness with LLMs

Innovation

• LLMs infer planning domains from single demo
• Automated domain construction without predefined info
• Integrates symbolic planning with motion skills