The Safety Dilemmas of Autonomous Driving

Abstract

We still do not see fully autonomous cars on our roads, although expectations were high a few years ago. We argue that this is at least partly due to the challenges to make them acceptably safe. These challenges are a lack of suitable standards for these highly complex systems, where functional safety is not sufficient. Additionally, the "safety dilemma" of artificial intelligence – the fact that artificial intelligence is needed but undesired by safety standards, poses a challenge. Security needs to be considered along with safety, as these two properties are highly interlinked.