Scale-Plan: Scalable Language-Enabled Task Planning for Heterogeneous Multi-Robot Teams

2026-03-09

AI Summary

This work addresses the challenges of long-horizon heterogeneous multi-robot task planning, including perceptual redundancy, reliance on handcrafted symbolic models, and the hallucination and weak environmental grounding of large language models (LLMs). To overcome these limitations, the authors propose Scale-Plan, a novel framework that integrates shallow LLMs with symbolic planning. Scale-Plan generates compact problem representations from natural language instructions, constructs action graphs using PDDL, and employs LLM-guided structured graph search to identify minimal relevant subsets of actions and objects. This enables efficient task decomposition, allocation, and long-horizon planning. A key innovation is the incorporation of a pre-filtering mechanism that enhances scalability and reliability. Evaluated on complex multi-agent tasks and the new MAT2-THOR benchmark, Scale-Plan consistently outperforms both pure LLM and hybrid LLM-PDDL baselines.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Multiagent Systems: Multiagent Planning
• Intelligent Robots: Motion and Path Planning

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information

Abstract

Long-horizon task planning for heterogeneous multi-robot systems is essential for deploying collaborative teams in real-world environments; yet, it remains challenging due to the large volume of perceptual information, much of which is irrelevant to task objectives and burdens planning. Traditional symbolic planners rely on manually constructed problem specifications, limiting scalability and adaptability, while recent large language model (LLM)-based approaches often suffer from hallucinations and weak grounding-i.e., poor alignment between generated plans and actual environmental objects and constraints-in object-rich settings. We present Scale-Plan, a scalable LLM-assisted framework that generates compact, task-relevant problem representations from natural language instructions. Given a PDDL domain specification, Scale-Plan constructs an action graph capturing domain structure and uses shallow LLM reasoning to guide a structured graph search that identifies a minimal subset of relevant actions and objects. By filtering irrelevant information prior to planning, Scale-Plan enables efficient decomposition, allocation, and long-horizon plan generation. We evaluate our approach on complex multi-agent tasks and introduce MAT2-THOR, a cleaned benchmark built on AI2-THOR for reliable evaluation of multi-robot planning systems. Scale-Plan outperforms pure LLM and hybrid LLM-PDDL baselines across all metrics, improving scalability and reliability.

Problem

• multi-robot task planning
• long-horizon planning
• large language models
• grounding
• scalability

Innovation

• LLM-assisted planning
• task-relevant abstraction
• action graph
• scalable multi-robot planning
• grounded reasoning