Can LLM Agents Solve Collaborative Tasks? A Study on Urgency-Aware Planning and Coordination

2025-08-20

AI Summary

This study investigates the structured collaboration capabilities of large language model (LLM)-based agents in multi-agent coordinated rescue tasks, focusing on task allocation, urgency awareness, and joint planning. We propose the Urgency-Aware Collaborative Planning Framework, evaluated in a graph-structured simulation environment that integrates embodied action modeling and a collaboration-sensitive evaluation suite—including task success rate, redundant action ratio, spatial conflict frequency, and weighted response efficiency. Results demonstrate that LLM agents effectively prioritize tasks, adaptively allocate roles based on environmental dynamics, and improve overall rescue timeliness. However, coordination failures persist in highly coupled scenarios involving resource contention and action synchronization. The work establishes a novel paradigm for benchmarking multi-agent LLMs, advances collaborative mechanism modeling, and provides empirical foundations for enhancing architectural robustness in distributed agent systems.

Technology Category

• Multiagent Systems: Coordination and Collaboration
• Machine Learning: Large Multimodal Models (LMMs)
• Planning, Routing, and Scheduling: Planning with Language Models

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
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

The ability to coordinate actions across multiple agents is critical for solving complex, real-world problems. Large Language Models (LLMs) have shown strong capabilities in communication, planning, and reasoning, raising the question of whether they can also support effective collaboration in multi-agent settings. In this work, we investigate the use of LLM agents to solve a structured victim rescue task that requires division of labor, prioritization, and cooperative planning. Agents operate in a fully known graph-based environment and must allocate resources to victims with varying needs and urgency levels. We systematically evaluate their performance using a suite of coordination-sensitive metrics, including task success rate, redundant actions, room conflicts, and urgency-weighted efficiency. This study offers new insights into the strengths and failure modes of LLMs in physically grounded multi-agent collaboration tasks, contributing to future benchmarks and architectural improvements.

Problem

• Evaluating LLM agents' collaborative task performance
• Assessing urgency-aware planning in multi-agent systems
• Investigating coordination failures in victim rescue scenarios

Innovation

• LLM agents for urgency-aware collaborative planning
• Graph-based environment with resource allocation strategies
• Coordination metrics evaluating task success and efficiency