Nearmap provides AI-driven peril scoring through the AI Features API, enabling property-level risk assessment for insurance underwriting, rating, and loss mitigation. By combining computer vision detections from high-resolution aerial imagery with geographical and historical data, these peril scores give insurers a transparent and actionable view of property vulnerability across multiple natural hazard types.
What are Perils?
Perils are natural hazard risk models that evaluate a property's vulnerability to specific types of weather-related damage. Each peril produces a vulnerability score graded from 1 (most vulnerable) to 5 (most resilient), trained on historical claims and damage datasets. These scores are derived from visible property attributes such as roof condition, roof shape, roof material, tree overhang, and defensible space.
Peril scores are available through the AI Features API and the Betterview platform.
How Peril Scores Work
Peril scoring combines two key components:
- Vulnerability: The likelihood that a specific structure will experience damage based on property condition. Nearmap's machine learning models analyze aerial imagery to identify risk factors such as missing shingles, roof staining, worn materials, yard debris, and vegetation encroachment.
- Hazard: The likelihood that a property will experience a peril event based on geographic and contextual factors. Hazard data is sourced from trusted third-party partners.
Together, these components provide a comprehensive risk picture that goes beyond regional hazard data alone by incorporating property-specific structural vulnerability.
Foundational Concept: Defensible Space
Defensible Space is a core component that feeds into multiple peril scores. It analyzes the area surrounding a building in three concentric zones (0–5 ft, 5–30 ft, and 30–100 ft), measuring the coverage of risk objects such as dense vegetation, yard debris, and adjacent structures within each zone. Defensible Space results are used as inputs to the Hurricane, Wildfire, and Wind vulnerability scores.
For full details on how Defensible Space zones are calculated, the risk objects analyzed, and coverage ratio formulas, see the Defensible Space - Overview article.
Peril types
Nearmap currently offers vulnerability scoring for the following peril types. Each article provides details on the specific property attributes evaluated, scoring methodology, available API fields, and regional availability.
Access and Availability
Peril scores are calculated in real time and may introduce some latency. As a default, they are excluded from the standard AI Features API response and must be explicitly requested. Refer to each peril's overview article for specific request parameters and response details.
Related Articles