Expand your Vision

18.02.2020

“Thinking of tomorrow today” – Hardly any other motto sums up the core concerns of the insurance business so succinctly. Of course, it does not apply exclusively to the insurance industry. If you want to be secure and successful tomorrow, you must take measures today. You need to get informed, network with others and make decisions. It is important to be open-minded, to embrace new ideas, and to keep an eye on the latest trends. And that is what we at ControlExpert practice in daily life. With one difference, however: We not only think about tomorrow today, but about the day after also. At the very least. After all, we have set ourselves an ambitious goal: We want motorists all over the world to have their claims settled fairly on the same day their vehicle was damaged. Granted, this may sound like science fiction.

But some of what we do today actually was science fiction until just recently. Up until a few years ago, the importance of the role that automatic image recognition would play in the current and future claim process could, at the very best, only be guessed at. A self-learning artificial intelligence that independently extracts, analyzes and qualifies information from documents and photographs of damaged vehicles? At that time a vision, today reality. Following intensive research, algorithm training, test runs and prototype development, the technology we have designed is now ready for the market and can be put to use. In a four-step process based on deep learning, an algorithm recognizes car parts on a photograph, distinguishes between damaged and undamaged parts, determines the severity of the damage and thus provides the information required to calculate the repair costs or provide an expert opinion.

The claims process, which still takes 27 days on average, can be shortened significantly by automated expert opinions. Although this is, of course, not enough to reduce a claims settlement from 27 days to one day, it is just one of the many building blocks that makes up our daily work here at ControlExpert. Our research department, for example, addresses all the parts of a claims process that take a long time and are inefficient. In addition to automatic image recognition, the focus lies on automating manual process steps and interpreting telematic data.

With their know-how, skills and curiosity, each of our 800 employees in 17 countries contributes not only to broadening our perspective as a company, but also to helping us attain our goals and exploit opportunities. In doing this, we rely on today’s technologies while at the same time continuing the search for the technologies of tomorrow. Or rather, of the day after tomorrow. Including advanced analytics, for example, the result of new classes of algorithms and data science that lead to new skills, such as transfer learning, which uses pre-trained machine learning models as advanced starting points for new technologies. Or digital ecosystems, post-classical compute and comms, edge analysis and generative adversarial networks. Topics such as these are already keeping us on our toes today – and bringing us closer to our vision.