An LLM-based multi-agent system for remote sensing analysis

Remote sensing analysis converts remotely sensed data into actionable insights across diverse fields. However, traditional data-driven approaches frequently overlook the technical challenges encountered by end-users lacking remote sensing expertise. To address this issue, we proposed a multi-agent system framework based on a large language model, designed to facilitate remote sensing analysis for non-experts. The framework consists of three main modules: Data, which includes remote sensing datasets and supplementary information; Tools, encompassing algorithms and visualization tools; and Brain, which provides AI-driven task management and reasoning. We employed a prototype system called ExpertsRS, developed on the AutoGen framework and DeepSeek-V3, to validate the proposed framework. This prototype features three LLM-powered agents with distinct roles and functionalities, collaborating to fulfill user requests. The system was evaluated through two experiments. The first experiment demonstrated that ExpertsRS could translate ambiguous queries into structured analytical workflows and produce outputs comparable to those of human experts. The second experiment indicated that ExpertsRS outperformed both a baseline LLM and a single-agent configuration in planning efficiency and result accuracy, with only a modest increase in token usage. Our system underscores the potential of large language model-based multi-agent system to assist end-users in overcoming technical barriers in remote sensing analysis.