Hazardweaver selects best scientific methods for analyzing disasters
HazardWeaver: Scientific Route Selection for Hazard Analysis Agents
Artificial Intelligence
Summary
Natural disasters like floods or earthquakes are tricky to understand because different scientific methods work better depending on the situation and available data. The authors created HazardWeaver, a system that figures out which scientific tools and data are best for studying a specific hazard as new information comes in. It can switch methods if conditions change, helping to produce reliable disaster analyses automatically. They also built a test set with many hazard scenarios and showed HazardWeaver works better than earlier systems.
What this means in practice
- •For emergency management teams: Use automated hazard analyses that select scientifically valid methods adapting as new data emerges to improve disaster response planning.
- •For environmental monitoring teams: Coordinate multiple scientific data sources and models through automated workflows that update analyses when new evidence arrives.
Authors
Wangshu Zhu, Xueqi Cheng, Liang Wu, Yushun Dong
Abstract
Understanding and assessing natural hazards is essential for disaster preparedness and risk reduction. Recent advances in large language models have spurred growing interest in AI agents for hazard analysis, particularly their ability to integrate scientific data, models, and tools into automated workflows. However, effective automation requires agents to determine which scientific methods are appropriate for a given event and executable with the available data and tools. As new evidence and execution results become available, these conditions can change, requiring agents to reconsider their choices. We formulate this problem as state-dependent scientific route selection and introduce HazardWeaver. Specifically, HazardWeaver first leverages the Hazard Knowledge Compiler to extract evidence-linked conditions governing scientific applicability, then its Hazard Capability Graph represents executable scientific capabilities and checks compatibility between their inputs and outputs. Using these complementary representations, the Hazard Weaver Agent component selects applicable and executable routes, carries out their workflows, and revises its decisions as the analysis state changes. To evaluate both the scientific outputs and the decisions that produce them, we introduce the Hazard Weaver Benchmark, comprising 141 instances across seven single-hazard domains and four multi-hazard interaction classes. The benchmark accommodates multiple valid scientific routes and evaluates output correctness, route validity, and justified abstention. Extensive experiments on this benchmark show that HazardWeaver outperforms existing agent systems, with the largest gains on tasks with multiple eligible scientific routes. Our code is publicly available at https://github.com/LabRAI/HazardWeaver.