Advanced Analytics in Insurance: Utilizing Building Footprints Derived from Machine Learning and High-Resolution Imagery

April 2025 

Authors 

Greg Dietzen, FCAS, MAAA
Consulting Actuary
Milliman, Inc.

Garrett Bradford, GISP
Principal and GIS Consultant
Milliman, Inc.

Kailey Adams
GIS Analyst
Milliman, Inc.

Claire Palmer, GISP
GIS Programmer
Milliman, Inc.

Description

Geospatial imagery and high-resolution image technology have spurred advances across many industries, including agriculture (Jain et al., 2017), transportation (Lee, Oh, & Ryu, 2003), disaster response (NOAA, 2022), and construction (Ma, Wang, Cheng, Li, & Tong, 2013). There are a wide variety of applications for the property and casualty (P&C) insurance industry as well, including underwriting, rating, and claims settlement.

One narrow but critical slice of geospatial data relevant to the P&C insurance industry is building location, typically inferred using address-based geocoding. Geocoding workflows typically result in a single point location that may or may not accurately represent the building(s) to be insured. New methods for determining building location, such as the use of structure footprints, may offer significant improvements in risk determination in certain cases.

The paper begins with an introduction to the use of remotely sensed imagery and artificial intelligence to delineate detailed building footprints across wide geographic scales. The addition of remotely sensed high-resolution elevation information (e.g., Light Detection and Ranging or LiDAR) to define key structural characteristics relevant to risk assessment will also be covered.

The paper then presents two case studies to illustrate the utility of building footprint data.

Material

Advanced Analytics in Insurance

Acknowledgments

The researchers’ deepest gratitude goes to those without whose efforts this project could not have come to fruition: the Project Oversight Group for their diligent work overseeing, reviewing and editing this report for accuracy and relevance.

 Project Oversight Group members:
Yuzhou Chen, PhD
Christopher Clickner, ACAS, MAAA
Marco De Virgilis, FCAS
Xinyi Hu
Min Ji, FSA, FIA, MAAA
Nicholas LaHaye, PhD
Hugo Lee, PhD
Cheok Yeow Leow, FSA
Daniel Lupton, FCAS, MAAA, CSPA
Daniel Pribe, FSA, MAAA
Priya Rohatgi, ASA

At the Society of Actuaries Research Institute:
Rob Montgomery ASA, MAAA, FLMI, Consultant - Research Project Manager

The Society of Actuaries Research Institute would like to acknowledge the generous contribution of the Casualty Actuarial Society to the funding of this research.

Questions or Comments?

Give us your feedback! Take a short survey on this report. Take Survey

If you have comments or questions, please send an email to research@soa.org