Objectives of the service
Wildfire is a growing threat across Southern Europe, yet insurance underwriters lack fast, affordable tools to assess risk at the level of individual properties or industrial sites. Today's options are slow and expensive: commissioning a physical inspection takes weeks and costs thousands of euros, while generic catastrophe models only give broad area-level estimates. Neither is practical for routine underwriting decisions.
HALO Dashboard solves this by delivering a structured wildfire risk score for any asset within minutes, using satellite data processed automatically in the cloud. Underwriters enter a location or upload a portfolio and receive a risk assessment covering vegetation hazard, proximity to fire-prone wildland areas, site defensibility, structural exposure, and current fire weather conditions. The service is accessed via a simple web interface or API, integrates into existing workflows without specialist training, and costs a fraction of a physical site visit. The goal is to make high-quality wildfire risk assessment fast enough and affordable enough to be used as a standard step in the insurance underwriting process — not just for exceptional cases.
Users and their needs
Target users are European property and casualty insurance professionals responsible for underwriting or risk managing large commercial and industrial assets in wildfire-prone regions — primarily Southern Europe, the Iberian Peninsula, Greece, and the Balkans.
Who they are: specialty property underwriters, catastrophe risk managers, and reinsurance analysts at mid-to-large insurance carriers and reinsurers operating across Italy, Spain, Portugal, Greece, and Germany.
Their needs and challenges:
- Need to assess wildfire exposure quickly before deciding whether to accept or price a risk, without waiting weeks for a physical inspection
- Require asset-level granularity — generic regional hazard maps are insufficient for individual industrial site decisions
- Work under combined ratio pressure that limits what they can spend on data per risk
- Must justify risk assessments to actuarial and compliance teams under Solvency II frameworks
- Currently have no standard satellite-based tool for wildfire — assessments are manual and inconsistent
During the study, underwriters and catastrophe risk managers from several European insurance organisations participated in structured workshops. Country focus: Greece, Germany, Spain, France, Switzerland.
Service/ system concept
HALO Dashboard delivers a per-asset wildfire risk score comprising five components: vegetation fire hazard (derived from satellite vegetation health data), wildland-urban interface exposure, site defensibility assessment, structural vulnerability classification, and fire weather risk. Results are presented as a structured report with an overall risk rating and component-level detail, accessible via a web dashboard or REST API.
How it works: The user submits an asset location (coordinates or address). The system automatically retrieves and processes the relevant satellite imagery and data products, runs the scoring algorithms, and returns a report — typically within minutes.
System architecture (simplified):
Satellite Data Sources
(Sentinel-2, Meteosat MTG)
↓
Data Ingestion &
Pre-processing Layer
↓
Scoring Engine
(Hazard · WUI · Defensibility
Vulnerability · Fire Weather)
↓
Risk Report Generator
↓
Web Dashboard / REST API
↓
Underwriter
All processing runs on cloud infrastructure. No local software installation is required. Historical data is retained to support portfolio-level queries and trend analysis.
Space Added Value
HALO Dashboard uses two primary satellite data sources:
Sentinel-2 (ESA Copernicus) provides multispectral optical imagery at 10–20 metre resolution. Vegetation health indices derived from these images (measuring greenness and moisture content) are used to classify fire hazard and identify areas of elevated fuel load — information that cannot be obtained reliably from ground-based or aerial sources at scale.
Meteosat Third Generation (MTG) (EUMETSAT) provides near-real-time fire detection and fire radiative power products, enabling dynamic fire weather risk assessment and active fire monitoring integrated into the risk score.
Added value over existing methods: Physical site inspections provide accurate assessments but cost EUR 2,000–8,000 per visit and take weeks to arrange — making them impractical for routine underwriting. Generic catastrophe models provide regional hazard estimates but cannot resolve individual asset risk. Satellite data enables HALO to assess any asset anywhere in the covered region within minutes, at a cost that is viable for commercial underwriting decisions. The combination of Sentinel-2's spatial resolution with MTG's temporal frequency is what makes both asset-level granularity and dynamic fire weather integration possible simultaneously.
Current Status
The ESA Kickstart phase concluded in June 2026 and the Final Review was held. The HALO processing pipeline — from satellite data ingestion through to API-delivered risk reports — has been built and technically validated. Core components covering vegetation hazard scoring, wildland-urban interface mapping, and defensible space analysis are operational.
On the commercial side, the study engaged thirteen European insurance organisations between March and May 2026 through a structured sequence of introduction calls, exploration sessions, asset-based workshops, and pricing discussions. Two full workshop sessions were held with one insurance company at catastrophe management level, during which HALO outputs were directly compared against their existing workflow on a live portfolio. A second insurer completed a conclusion call with a follow-up scheduled for Q3 2026.
Key finding: the service is viable for large industrial and commercial assets; the residential property market is not a target. A qualified commercial pipeline has been established. It is is planned to apply for a follow on Pilot Project phase, with the objective of converting the leading engagement into a first paid pilot agreement.
