Munich Re is convinced that increasing automation will solve many problems in the area of claims settlement. The aim is to progressively simplify and accelerate processes that are mostly manual today, with the help of the latest technologies. This could massively reduce the costs of claims settlement. But insurance customers would also benefit from faster indemnification, which in turn would have a positive knock-on effect for primary insurers in the form of higher levels of customer satisfaction.

The key to such improvements lies in the consistent use of remote-sensing systems: aerial photos help to define the areas affected by a natural catastrophe, and to identify and classify losses. And if a company has a geocoded portfolio, the exposure can then be calculated with minimum effort. This is already the state of the art today. In cases where satellite images are not available, or not in good enough quality, what are known as HALE drones could be used instead. This acronym stands for High Altitude Long Endurance. High altitude in this context means deployment at more than 15 km above the surface of the Earth. Long endurance means that the drones can operate for up to three months with the help of solar-powered drives. The HALE drones provide images with a higher resolution than satellite photographs, and these images are now more widely available with the increasing spread of the technology.

If claims are classified according to their severity, loss adjusters can be deployed more efficiently, allowing specialists to be sent to investigate the most complex cases. Munich Re anticipates that automation will have reached a high level of sophistication by 2025, driven by remote sensing, geoinformation systems, image-processing algorithms and digital platforms. Satellites alone are not enough, because their image resolution is too coarse – for example to identify missing tiles on a roof. Alternatively, the policyholder could take pictures or make videos of the damaged property and send these to the insurer on a mobile phone. In conjunction with aerial image analysis, this could help create an accurate damage profile.

A loss estimation engine will then be able to calculate the nature and amount of damage. Data from remote sensing and from the policyholder will be checked using algorithms and supplemented with experience values from the past. The degree of automation will increase further with the spread of sensors that can recognise, for example, whether a wall has shifted or sunk. In terms of specific applications, the focus will initially be on the natural hazards of storms and earthquakes, because the damage they cause is the easiest to identify using remote-sensing methods. While the assessment will initially be carried out by a claims technician, it is conceivable that, with the growing proliferation of artificial intelligence, an algorithm will be able to assume this task.