AXA Fast

Proof of Concept

Project started with a local start up company at end of 2019 with only front and rear bumpers in scope as a PoC, with the aim of correctly identifying damaged parts and to give accurate repair / replace decisions. A local start up partner processed approximately 500 K photos and claim informations from our historical data and an artificial intelligence model is developed. It was ensured that the artificial intelligence recognized the bumper part of the vehicle in the first place and gave correct result about the claim severity.

Image Capturing via Artifical Intelligence Assistance

The tool make an assessment whether 2 and more spare parts is damage with artificial intelligence integration by adding a final step at the FNOL. If the result is fast track according to whole rule set and the tool decision is 2 and less than 2 spare part is damaged the claim operated fast track.

In order to improve image quality, andhence model accuracy, a mobile app was also developed that guides user to take photos from the correct angles and distances, and at the desired light conditions. The app also includes ghost images suitable for vehicle types to help users with the photo capturing process