Abstract
The automated and intelligent processing of massive remote sensing (RS) datasets is critical in Earth observation (EO). Existing automated systems are normally task-specific, lacking a unified framework to manage diverse, end-to-end workflows--from data preprocessing to advanced interpretation--across diverse RS applications. To address this gap, this paper introduces Cangling-KnowFlow, a unified intelligent agent framework that integrates a Procedural Knowledge Base (PKB), Dynamic Workflow Adjustment, and an Evolutionary Memory Module. The PKB, comprising 1,008 expert-validated workflow cases across 162 practical RS tasks, guides planning and substantially reduces hallucinations common in general-purpose agents. During runtime failures, the Dynamic Workflow Adjustment autonomously diagnoses and replans recovery strategies, while the Evolutionary Memory Module continuously learns from these events, iteratively enhancing the agent’s knowledge and performance. We evaluated Cangling-KnowFlow on the KnowFlow-Bench, a novel benchmark of 324 workflows inspired by real-world applications, testing its performance across 13 top LLM backbones from open-source to commercial. Across all complex tasks, Cangling-KnowFlow surpassed the Reflexion baseline by at least 4% in Task Success Rate. As the first comprehensive validation along this emerging field, this work demonstrates the great potential of Cangling-KnowFlow as a robust, efficient, and scalable automated solution for complex EO challenges by leveraging expert knowledge (Knowledge) into adaptive and verifiable procedures (Flow).
Framework: Orchestrator Agent + Procedural Knowledge Base + Dynamic Execution Engine + Evolutionary Memory Module.
Method
Workflow Tools: curated tool library, workflow templates, and parameterized Directed Acyclic Graphs.
Dynamic Adjustment: runtime monitoring, failure detection, and graph-level repair to keep long-horizon execution on track.
Task Support: 162 simple & complex tasks for diverse Earth observation scenarios.
Results
KnowFlow-Bench: across 13 LLMs, Cangling-KnowFlow boosts Task Success Rate/First-Pass Accuracy and cuts Number of Tool Calls by ~38% on complex remote sensing tasks.
Transfer on ThinkGeo benchmark: significantly outperforms SOTA agents, showing cross-toolset adaptability.
Ablation: Workflow Library is the largest contributor; Dynamic Adjustment and Learning Capability add robustness and evolution.
BibTeX
@article{CangLingKnowFlow2025,
title={CangLing-KnowFlow: A Unified Knowledge-and-Flow-fused Agent for Comprehensive Remote Sensing Applications},
author={Chen, Zhengchao and Wang, Haoran and Yao, Jing and Ghamisi, Pedram and Zhou, Jun and Atkinson, Peter M and Zhang, Bing},
journal={arXiv preprint arXiv:2512.15231},
year={2025}
}