The company fully digitalises the insurance value chain through proprietary technology platforms and advanced data analytics. Its infrastructure, risk analysis capabilities, and full-stack digital approach enable individual policyholders across Europe to access customisable insurance with greater affordability and improved service quality. Its sophisticated insurance models and ability to translate analytics into commercial performance have reshaped the insurance market in Italy, supporting rapid expansion into Spain and the UK.
While its risk models were designed to better understand and predict human behaviour, they did not fully capture the physical environment in which risk actually unfolds. Yet road risk is shaped not only by who the driver is, but by where driving occurs — including terrain, road geometry, micro-climate, and surrounding environmental conditions. These spatial factors are rarely included in traditional datasets, despite their daily influence on road accidents.
The insurer identified a structural limitation common across the industry. While behavioural and historical datasets are rich, geospatial context often remains fragmented, static, or difficult to scale. That’s where Earthpulse came in.
Key limitations included:
Limited access to harmonised, scalable geospatial indicators
Fragmented data sources that were difficult to integrate into existing models
The need to evaluate road quality and safety consistently across regions
Increasing pressure to support expansion into new markets without slowing model development
Earthpulse collaborated with the insurer to generate a structured, large-scale, automated set of geospatial indicators describing road environments across Italy, Spain, and the UK, covering over 1 million km².
Using SPAI, our Satellite Processing by Artificial Intelligence platform, we integrated multiple layers of information — from climate records and topography to road morphology, environmental conditions, and contextual datasets. This enabled the creation of a comprehensive geospatial catalogue designed for direct integration into risk models.
The indicators combined satellite data and advanced geospatial analysis across multiple domains, including:
road density and connectivity
slope and micro-topography
population and surrounding amenities
environmental and weather patterns
All indicators were produced in a continuous, automated, and fully scalable manner, ensuring consistency across regions and countries.
During the first phase, the insurer selected representative regions across multiple countries to assess the impact of geospatial enrichment on road-risk modelling.
Earthpulse analysed a broad catalogue of geospatial features spanning road structure, connectivity, terrain, population context, surrounding amenities, and weather conditions.
Both teams jointly evaluated each feature for statistical relevance and business impact, ensuring that only the most meaningful spatial signals were selected for operational use.
Road risk is shaped by micro-conditions that traditional datasets often overlook.
For example, a north-facing road may remain colder and damper, increasing ice risk. Micro-topography influences braking distance and accident severity. Local climate patterns can create fog corridors and shadow zones. Road curves, slopes, surface exposure, and surrounding context quietly shape risk every day.
These factors are rarely visible from the ground — but they are consistently visible from space.
The first phase confirmed that adding structured geospatial context can materially improve risk modelling, even within already advanced analytical systems.
Through joint validation, a subset of high-impact geospatial indicators was selected as the foundation for operational deployment. The scope was refined to focus on indicators delivering the greatest predictive value, and rollout strategy aligned with business priorities.
This validation enabled both teams to:
focus on indicators that add measurable predictive value
refine the operational rollout scope
align investment with demonstrated business impact
With Earthpulse’s SPAI platform, this level of detail can now be generated automatically at national scale and integrated directly into risk-prevention workflows.
The validated geospatial dataset generated by Earthpulse provides:
A scalable platform for accessing road and traffic intelligence
Faster integration of satellite and geospatial data into core models
Improved risk differentiation across regions
A future-proof foundation to support expansion into new markets