Bookmarks

Bluesky and University of Leicester Bring AI-Driven Geospatial Intelligence to Insurance

avatar
Michael Johnson
post-picture

UK-based aerial mapping specialist Bluesky International has partnered with the University of Leicester on a research initiative designed to apply artificial intelligence and machine learning to geospatial data products for the insurance industry. Supported by funding from the SPRINT programme, the project explores how satellite imagery and aerial survey data can be transformed into automated, high-confidence intelligence about buildings and urban environments.

The collaboration focuses on improving how insurers assess structural characteristics and environmental factors at scale. By combining AI techniques with remote sensing data, the project aims to automatically identify building types, construction features, and other attributes relevant to risk evaluation and underwriting.

Applying AI to Urban Geospatial Data

According to Bluesky’s Technical Director, James Eddy, the company was keen to investigate how modern machine learning methods could streamline data classification for insurance applications. With previous experience working alongside academic institutions, Bluesky approached the University of Leicester to access specialist expertise in mathematical modelling and data science.

The project was aligned with the SPRINT (Space Research & Innovation Network for Technology) programme due to its emphasis on space-derived data. This enabled researchers to analyse information in multiple forms and investigate techniques that could ultimately be commercialised for use not only in insurance, but across a wider range of industries.

Academic Expertise Meets Commercial Data

Professor Ivan Tyukin, Professor of Applied Mathematics at the University of Leicester, highlighted the growing importance of advanced analytics as satellite data volumes continue to expand rapidly. The research team is applying state-of-the-art machine learning approaches to extract meaningful insights from satellite datasets and integrate them with Bluesky’s proprietary aerial data.

This fusion of academic research and commercial datasets is intended to significantly enhance Bluesky’s analytical capabilities, delivering more reliable and scalable intelligence services.

From Imagery to Actionable Intelligence

Bluesky brings extensive experience in national-scale geospatial data production, including high-resolution aerial photography and 3D height models. Working with specialists in object classification and remote sensing, the project explores how visible and infrared imagery—used individually or in combination—can support robust automated analysis.

The goal is to generate accurate, repeatable outputs that insurers can rely on when assessing exposure, property characteristics, and environmental risk factors.

Funding and Programme Support

The research is funded through a grant from the £4.8 million SPRINT programme, which provides businesses with access to university expertise in space data and technology. SPRINT is designed to help organisations translate academic research into commercially viable products and services.

About the Partners

Bluesky International is one of the UK’s leading providers of aerial survey and geospatial data, maintaining nationwide coverage of digital aerial imagery and elevation models across Great Britain and the Republic of Ireland. The company also delivers bespoke LiDAR and imaging surveys for both public and private sector clients, and is known for innovative products such as the National Tree Map.

The University of Leicester is internationally recognised for research excellence and innovation, with strong expertise across mathematics, space science, and data analytics. Through initiatives such as SPRINT, the university supports collaboration between academia and industry to unlock the commercial value of space-enabled data and technologies.

Read more

post-picture
post-picture
post-picture
post-picture
post-picture
post-picture
post-picture
post-picture
post-picture
post-picture
post-picture
post-picture
post-picture
post-picture
post-picture