mlsanityData to 14 Oct 2026

GeoDisaster: Benchmarking Orchestrated Agents for Operational Disaster Geo-Intelligence

arXiv 2606.17246 ↗First posted 15 Jun 2026cs.CV cs.MAcited by 0

On the map: LLM reasoning → Spatial and embodied reasoning

also in 1 more
Abstract (arXiv)

Remote-sensing vision-language models (RS-VLMs) have advanced Earth-observation analysis toward visual interpretation and instruction-following, yet fall short of operational geo-intelligence, which demands tool-grounded spatial reasoning and structured, evidence-backed decisions. We introduce GeoDisaster, an operational geospatial disaster reasoning benchmark with 2,921 verified instances across 43 question types and five task families: deforestation monitoring, multi-hazard analysis, building-damage assessment, flood-safe routing, and Sentinel-1 SAR flood monitoring. Instances integrate heterogeneous EO/GIS evidence-optical and SAR imagery, raster masks, vector geometries, road networks, and exposure layers-spanning hazard detection, damage assessment, exposure estimation, and diagnostic report generation. Ground-truth answers are grounded in executable geospatial workflows and deterministic consistency checks, removing the need for language-model annotation. We further propose an orchestrated multi-agent framework with 18 disaster-oriented tools, where role-specialized agents coordinate through explicit execution contracts, aligned via Role-Contract Expectation Alignment (RCEA): failure-aware supervised fine-tuning combined with contract-grounded reinforcement learning over dense step-level signals. Experiments show that GeoDisaster challenges existing RS-VLMs and agentic systems, while RCEA improves tool use, evidence grounding, state consistency, and decision generation.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 36 of its 56 references that are in the map. First posted on arXiv 15 Jun 2026.

Builds on

36 of 56 references.

Closest ancestors · 4

Same direction, newest first

  1. 2026OpenEarthAgent: A Unified Framework for Tool-Augmented Geospatial Agents19 Feb
  2. 2026GeoReason: Aligning Thinking And Answering In Remote Sensing Vision-Language Models Via Logical Consistency Reinforcement Learning7 Jan
  3. 2025Empowering LLM Agents with Geospatial Awareness: Toward Grounded Reasoning for Wildfire Response14 Oct
  4. 2025ThinkGeo: Evaluating Tool-Augmented Agents for Remote Sensing Tasks29 May

Other references · 32

  1. 2022Training language models to follow instructions with human feedbackcited by 15,971
  2. 2017Proximal Policy Optimization Algorithmscited by 15,215
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  4. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  5. 2019Fine-Tuning Language Models from Human Preferencescited by 2,150
  6. 2023ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIscited by 1,947
  7. 2025Why Do Multi-Agent LLM Systems Fail?cited by 604
  8. 2025Group-in-Group Policy Optimization for LLM Agent Trainingcited by 497
  9. 2023AgentTuning: Enabling Generalized Agent Abilities for LLMscited by 338
  10. 2023GeoChat: Grounded Large Vision-Language Model for Remote Sensingcited by 328
  11. 2024EarthGPT: A Universal Multi-modal Large Language Model for Multi-sensor Image Comprehension in Remote Sensing Domaincited by 184
  12. 2020FloodNet: A High Resolution Aerial Imagery Dataset for Post Flood Scene Understandingcited by 181
20 more
  1. 2019xBD: A Dataset for Assessing Building Damage from Satellite Imagerycited by 139
  2. 2016OSMnx: New Methods for Acquiring, Constructing, Analyzing, and Visualizing Complex Street Networkscited by 110
  3. 2024Change-Agent: Towards Interactive Comprehensive Remote Sensing Change Interpretation and Analysiscited by 63
  4. 2024RS-Agent: Automating Remote Sensing Tasks through Intelligent Agentcited by 49
  5. 2025LLM Collaboration With Multi-Agent Reinforcement Learningcited by 49
  6. 2025Earth-Agent: Unlocking the Full Landscape of Earth Observation with Agentscited by 46
  7. 2022RescueNet: A High Resolution UAV Semantic Segmentation Benchmark Dataset for Natural Disaster Damage Assessmentcited by 40
  8. 2025DisasterM3: A Remote Sensing Vision-Language Dataset for Disaster Damage Assessment and Responsecited by 36
  9. 2025Multi-Agent Geospatial Copilots for Remote Sensing Workflowscited by 22
  10. 2023HRVQA: A Visual Question Answering Benchmark for High-Resolution Aerial Imagescited by 21
  11. 2025Multi-Agent Deep Research: Training Multi-Agent Systems with M-GRPOcited by 18
  12. 2024Evaluating Tool-Augmented Agents in Remote Sensing Platformscited by 17
  13. 2025RSVLM-QA: A Benchmark Dataset for Remote Sensing Vision Language Model-based Question Answeringcited by 17
  14. 2026Dr. MAS: Stable Reinforcement Learning for Multi-Agent LLM Systemscited by 16
  15. 2021VQA-Aid: Visual Question Answering for Post-Disaster Damage Assessment and Analysiscited by 15
  16. 2024RescueADI: Adaptive Disaster Interpretation in Remote Sensing Images with Autonomous Agentscited by 11
  17. 2025CORE: Full-Path Evaluation of LLM Agents Beyond Final Statecited by 9
  18. 2026SELAUR: Self Evolving LLM Agent via Uncertainty-aware Rewardscited by 9
  19. 2025End-to-End Optimization of LLM-Driven Multi-Agent Search Systems via Heterogeneous-Group-Based Reinforcement Learningcited by 8
  20. 2026GeoMMBench and GeoMMAgent: Toward Expert-Level Multimodal Intelligence in Geoscience and Remote Sensingcited by 6

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Can LLM Agents Respond to Disasters? Benchmarking Heterogeneous Geospatial Reasoning in Emergency Operations
  2. 2026TerraLogic: A Benchmark for Hierarchical Geospatial Reasoning in Earth Observation
  3. 2026Decoding the Disaster: Multi-Task Geospatial Reasoning with Vision-Language Models and Crowdsourced Imagery for Disaster Mapping
  4. 2026Obshazard-bench: Benchmarking Multimodal Foundation Models for Real-Time Disaster Intelligence from Raw Earth Observation Streams
  5. 2026DisasterInsight: A Building-Centric Benchmark for Evaluating Vision--Language Models in Disaster Response
  6. 2026Agentic AI for Remote Sensing: Technical Challenges and Research Directions
  7. 2026Risk-Aware LLM Agents for Geospatial Data Retrieval: Design and Preliminary Adversarial Evaluation
  8. 2026Beyond detection: cooperative multi-agent reasoning for rapid onboard EO crisis response