AI Case Study
TensorFlight assists with commercial property insurance evaluation using computer vision to investigate building properties
"TensorFlight automatically analyzes satellite, aerial, drone and street view imagery in order to automate slow and expensive manual in-person property inspections."
Industry
Financial Services
Real Estate Development Operations
Project Overview
"TensorFlight builds a database of all buildings in the world, providing features like building footprint, number of stories, construction type or square footage. Our data is focused on commercial property insurance, but it is also relevant for other industries". Current capabilities include windborne debris,
mechanical equipment on roof, commercial, residential and temporary buildings footprints with 80% accuracy; vehicles, parking spaces, dead and live trees with 95% accuracy.
Reported Results
Potential benefits include according to the vendor include:
"Underwriting: Reduce premium leakage by highlighting areas to investigate for an underwriter - e.g. potential building degradation.
Reinsurance: No longer get hindered by agents, brokers or property owners sending you faulty information. Better understand catastrophic risk by understanding factors like number of stories and construction type. Risk: Better understand exposure of your portfolio. Proactively monitor changes to the property, even after the policy is sold."
Technology
Function
Operations
General Operations
Background
Conducting building inspections using a variety of image data taken from sources such as drones and satellites allows for more frequent updates than a manual inspection.
Benefits
Data
Satellite, aerial, drone and street view imagery