LaMMA-P: Generalizable Multi-Agent Long-Horizon Task Allocation and Planning with LM-Driven PDDL Planner

2024-09-30

arXiv.org

AI Summary

To address the challenges of subtask identification and dynamic allocation for long-horizon tasks in heterogeneous robotic collaboration, this paper proposes the first large language model (LLM)-driven multi-agent PDDL planning framework, enabling end-to-end mapping from natural language instructions to symbolic multi-agent plans. Our method integrates LLM-based semantic understanding, hierarchical task decomposition, heuristic symbolic search, and a dynamic load-balancing mechanism. We further introduce MAT-THOR—the first long-horizon, multi-agent benchmark tailored to household environments, featuring dual difficulty levels. Experiments demonstrate that our approach achieves a 105% improvement in task success rate and a 36% increase in average planning efficiency over MAT-THOR, significantly outperforming existing LLM-based multi-agent planners. Moreover, it exhibits strong cross-task generalization capability.

Technology Category

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

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• 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: Multilingual and cross-lingual Web search

Abstract

Language models (LMs) possess a strong capability to comprehend natural language, making them effective in translating human instructions into detailed plans for simple robot tasks. Nevertheless, it remains a significant challenge to handle long-horizon tasks, especially in subtask identification and allocation for cooperative heterogeneous robot teams. To address this issue, we propose a Language Model-Driven Multi-Agent PDDL Planner (LaMMA-P), a novel multi-agent task planning framework that achieves state-of-the-art performance on long-horizon tasks. LaMMA-P integrates the strengths of the LMs' reasoning capability and the traditional heuristic search planner to achieve a high success rate and efficiency while demonstrating strong generalization across tasks. Additionally, we create MAT-THOR, a comprehensive benchmark that features household tasks with two different levels of complexity based on the AI2-THOR environment. The experimental results demonstrate that LaMMA-P achieves a 105% higher success rate and 36% higher efficiency than existing LM-based multiagent planners. The experimental videos, code, datasets, and detailed prompts used in each module can be found on the project website: https://lamma-p.github.io.

Problem

• Handles long-horizon task allocation for robot teams.
• Improves subtask identification and allocation efficiency.
• Enhances generalization across complex household tasks.

Innovation

• Language Model-Driven PDDL Planner
• Multi-Agent Task Planning Framework
• Integration of LMs and Heuristic Search