Products / TwinCar


TwinCar Concept

When you are claiming insurance to an insurance company, after selecting a suitable picture from a number of registered photos, it recognizes the parts of the car from the pictures and identifies the type of damage seen on the pictured parts. It is an automatic repair cost estimation system that estimates how much damage has occurred and calculates the estimate.

Auto estimate service based on image recognition of damaged cars  

Provides Depth Measurement and Estimation Extraction Algorithm for Vehicle Damage Area.

Provides instant repair cost estimation of damaged cars

Adapted by the Korea Insurance Development Institute, an integrated control agency for insurance companies in Korea.

Image recognition process of damaged cars

Filter model

When claiming for insurance, in addition to car photos, various pictures such as ID cards and other required documents are imported as well. It is a model to classify only the car photos that are the target of recognization among various such photos.

Also, it separates images of personal information such as ID and phone numbers for the management for privacy data later on.

Part recognition model

It is a model to accurately determine the parts of a car within various pre-registered images.

Damage Type model

This model is for detecting the type and area of damage. The final damage depth is determined by comprehensively considering the type and area of damage.


AgileSoDA supports companies to transform AI capabilities and output into internal intellectual assets and to resolve data quality problems. It also provides algorithm-based technologies and services including reinforcement learnings for optimal business decision.

Contact US