MapAgent: A Hierarchical Agent for Geospatial Reasoning with Dynamic Map Tool Integration

2025-09-07

AI Summary

Current AI agents exhibit limited proficiency in geospatial tasks—such as spatial reasoning, multi-step planning, and real-time map interaction. To address this, we propose GeoAgent: a hierarchical multi-agent framework that decouples high-level planning from low-level execution and introduces a dedicated map-tool agent to enable parallel API scheduling and collaborative tool orchestration. This design reduces cognitive load, improves tool selection accuracy, and enhances cross-API coordination fidelity. GeoAgent integrates dynamic tool invocation, fine-grained task decomposition, and a customized geospatial tool suite, all embedded within a lightweight agent scaffolding architecture. Evaluated on four geospatial benchmarks, GeoAgent consistently outperforms state-of-the-art tool-augmented and agent-based baselines. The implementation is publicly available.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Computer Vision: Remote Sensing / Geospatial AI
• Multiagent Systems: Multiagent Planning

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Agentic AI has significantly extended the capabilities of large language models (LLMs) by enabling complex reasoning and tool use. However, most existing frameworks are tailored to domains such as mathematics, coding, or web automation, and fall short on geospatial tasks that require spatial reasoning, multi-hop planning, and real-time map interaction. To address these challenges, we introduce MapAgent, a hierarchical multi-agent plug-and-play framework with customized toolsets and agentic scaffolds for map-integrated geospatial reasoning. Unlike existing flat agent-based approaches that treat tools uniformly-often overwhelming the LLM when handling similar but subtly different geospatial APIs-MapAgent decouples planning from execution. A high-level planner decomposes complex queries into subgoals, which are routed to specialized modules. For tool-heavy modules-such as map-based services-we then design a dedicated map-tool agent that efficiently orchestrates related APIs adaptively in parallel to effectively fetch geospatial data relevant for the query, while simpler modules (e.g., solution generation or answer extraction) operate without additional agent overhead. This hierarchical design reduces cognitive load, improves tool selection accuracy, and enables precise coordination across similar APIs. We evaluate MapAgent on four diverse geospatial benchmarks-MapEval-Textual, MapEval-API, MapEval-Visual, and MapQA-and demonstrate substantial gains over state-of-the-art tool-augmented and agentic baselines. We open-source our framwork at https://github.com/Hasebul/MapAgent.

Problem

• Addressing geospatial tasks requiring spatial reasoning and map interaction
• Overcoming limitations of flat agent frameworks for geospatial APIs
• Reducing cognitive load and improving tool selection accuracy

Innovation

• Hierarchical multi-agent framework for geospatial reasoning
• Decouples planning from execution with specialized modules
• Adaptively orchestrates map APIs in parallel for data fetching