AI summaryⓘ
The authors identify that traditional methods struggle to handle the huge and complex Earth observation data now available, leading to the need for smarter Remote Sensing (RS) agents. They introduce SimCRAFT, a framework that helps create a smaller, more efficient model (7 billion parameters) specialized for RS tasks, which avoids relying on large general-purpose language models. They built a large dataset called SimRS-14k using a system that ensures workflows are correct and compatible with sensors and tools, and developed a training method called CRAFT to help the model apply existing procedures to new problems. Their experiments show that SimCRAFT performs very well compared to other large language models and specialized agents, making it a useful option for cases with limited computing resources.
Remote Sensing (RS)Large Language Models (LLMs)SimCRAFTWorkflow PlanningMultiagent SystemsDataset SynthesisFine-TuningContextual RetrievalStandard Operating Procedures (SOPs)Resource-Constrained Deployment
Authors
Haoran Wang, Jing Yao, Xu Yang, Zeqing Wang, Yang Zhang, Pedram Ghamisi, Zhengchao Chen
Abstract
The unprecedented surge in Earth observation data volume and diversity has exposed a critical bottleneck for traditional manual workflows, catalyzing the emergence of Remote Sensing (RS) Agents. However, the practical deployment of these advanced agents is severely hindered by their heavy reliance on large-scale general-purpose LLMs, which lack deep domain expertise and impose prohibitive infrastructure demands. To resolve this, we propose SimCRAFT, a model-agnostic framework that distills sophisticated RS orchestration capabilities into a compact 7B-scale model. Addressing data scarcity, we first pair a multiagent synthesis engine with a Mock Execution Engine that checks schema correctness, inter-tool dependencies, and sensor/tool compatibility, producing SimRS-14k, a large-scale, constraint-validated workflow planning corpus. Second, we propose Contextual Retrieval-Augmented Fine-Tuning (CRAFT) that finetunes the model to reason analogically by adapting retrieved Standard Operating Procedures to novel queries under a noise-robust objective, generalizing RAFT to multi-step RS workflow planning without mechanical copying. Extensive experiments demonstrate that SimCRAFT-7B significantly outperforms openweights LLMs and rivals advanced closedsource models and specialized RS agents, while reproducing across three 7B backbones. This work contributes a competitive open-weights baseline for lightweight RS intelligence, enabling efficient autonomous deployment under resource-constrained or resource-conserving conditions.