Experience-Driven Multi-Agent Systems Are Training-free Context-aware Earth Observers

2026-01-30

AI Summary

This work addresses the frequent failure of large language model (LLM) agents in tool-intensive, multimodal, and long-horizon Earth observation tasks, often caused by suboptimal tool parameter configuration and inadequate error recovery. To overcome these limitations, the authors propose GeoEvolver, a training-free, self-evolving multi-agent system that acquires domain knowledge through structured interactions, decomposes complex tasks into subgoals, and explores diverse tool parameter configurations at the subgoal level. GeoEvolver incorporates an evolvable dynamic memory mechanism that continuously enhances tool-specific expertise by integrating root-cause analysis of failures with distilled contextual examples. Evaluated across three Earth observation benchmarks, GeoEvolver consistently improves end-to-end task success rates by an average of 12% over multiple LLM backbones.

Technology Category

• Machine Learning: Evolutionary Learning
• Multiagent Systems: Multiagent Learning
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• 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

Abstract

Recent advances have enabled large language model (LLM) agents to solve complex tasks by orchestrating external tools. However, these agents often struggle in specialized, tool-intensive domains that demand long-horizon execution, tight coordination across modalities, and strict adherence to implicit tool constraints. Earth Observation (EO) tasks exemplify this challenge due to the multi-modal and multi-temporal data inputs, as well as the requirements of geo-knowledge constraints (spectrum library, spatial reasoning, etc): many high-level plans can be derailed by subtle execution errors that propagate through a pipeline and invalidate final results. A core difficulty is that existing agents lack a mechanism to learn fine-grained, tool-level expertise from interaction. Without such expertise, they cannot reliably configure tool parameters or recover from mid-execution failures, limiting their effectiveness in complex EO workflows. To address this, we introduce \textbf{GeoEvolver}, a self-evolving multi-agent system~(MAS) that enables LLM agents to acquire EO expertise through structured interaction without any parameter updates. GeoEvolver decomposes each query into independent sub-goals via a retrieval-augmented multi-agent orchestrator, then explores diverse tool-parameter configurations at the sub-goal level. Successful patterns and root-cause attribution from failures are then distilled in an evolving memory bank that provides in-context demonstrations for future queries. Experiments on three tool-integrated EO benchmarks show that GeoEvolver consistently improves end-to-end task success, with an average gain of 12\% across multiple LLM backbones, demonstrating that EO expertise can emerge progressively from efficient, fine-grained interactions with the environment.

Problem

• Earth Observation
• Multi-Agent Systems
• Tool-intensive Tasks
• Long-horizon Execution
• Geo-knowledge Constraints

Innovation

• experience-driven
• training-free
• multi-agent system
• context-aware
• tool-parameter optimization