Skip to main content

The cooperation between claimbird and the TNT aims to meet the requirements for automated damage detection. For this purpose, a data set is built up, trained for a deep learning system, which can support the insurance companies in analyzing and evaluating claims.  The solution offers insurers the possibility of cost-saving and time-efficient regulation, particularly of small claims.


The number of new registrations for cars in Germany has increased by over 30% in the last 10 years. The growing number of road users leads to an increasing number of damage cases that are reported to insurance companies. In most cases, insurers must decide whether the use of an expert is necessary. The appraisals often cost several hundred euros and thus represent a not inconsiderable cost block in the claims settlement process. The decision if an expert is needed requires the use of trained employees who analyze the data and images of each claim in detail.

Automated image processing has the potential to simplify this task considerably. The reason for this is the use of so-called “deep learning systems”, which learn abstract representations based on data in order to solve a wide variety of tasks.

At the Institute for Information Processing, Professor Rosenhahn and his team are researching new methods for automated image analysis. These are used to recognize objects and their properties in pictures. This makes it possible to identify and classify damage to car components.

The startup claimbird is a spin-off from Leibniz Universität Hannover and with claimBuddy it develops an intelligent solution for the claims process at the interface between insurance companies, customers, experts and workshops.

Abbildung Schadenregulierung