Continuous, Systematic Risk Mapping of Roads as an Input for Dynamic Risk Management (DRM) in Autonomous Systems

Abstract

The risk potential on German roads remains high: Even in 2021 with less traffic due to the Covid-pandemic, the police counted 2.3 million traffic accidents [1]. Many accidents occur due to individual mistakes of road users. Dangerous situations are often misjudged or not recognized on time, for example, due to distraction while driving [2]. Autonomous systems in vehicles show the potential to avoid driver-related accidents, but for a Dynamic Risk Management (DRM) reliable data is needed. This is exactly where the project “Early Detection of Dangerous Areas in road traffic using smart data - EDDA+” comes in. The road hazard map created by using the EDDA+ method evaluates the Germany-wide road network according to a hazard score. This digital, safety-related data includes a lot of contextual information like weather or daytime conditions and can be used as an additional basis for the DRM risk analysis. For example, an autonomous system could react more sensitively at road areas where the hazard score is high. This continuously updated hazard map is published on www.gefahrenstellen.de and also available in a more detailed way on our platform for professional users such as local authorities, police, science, engineering offices, navigation providers and car manufacturers.