Introduction

Road transport and the automotive industry are currently undergoing a period of radical technological change. Noticeable trends towards electric vehicles, networking and automation are the dominant themes. The rise in digitalisation such changes have brought with them has focussed attention on the question of IT security in vehicles. In order to minimise risks to the welfare of road users, all aspects of IT security must be considered.

A variety of wireless interfaces and sensors in modern vehicles provide attackers with possible points of entry. For example, vehicles of the future will be able to exchange messages about what is happening on the roads as part of cooperative intelligent transport systems. Data will also be sent to vehicles from the road transport infrastructure. Traffic lights and other road fixtures will therefore also become possible targets for attacks.

At the same time, advances in automation are reducing the amount of human intervention that is required when driving. The key player in this context is artificial intelligence (AI). A scale made up of five levels is used to classify the extent to which a vehicle is automated, from support provided by assistance systems through to fully automated driving in any and all driving situations. AI methods are used in particular to process input data from optical sensors such as cameras. Such data can be used to classify road signs or detect the presence of pedestrians, for example, and control the response of the vehicle accordingly.

However, the potential of AI in automation is not immune to risks either. First, AI systems are vulnerable to new types of attack which can have serious consequences at any stage of their lifecycle. Second, AI systems must be able to continue to function reliably under all manner of ambient conditions without their automation being compromised. Among other things, the BSI is working to find ways to improve the resilience of AI systems to both attacks and changes in ambient conditions, and to assess this resilience in a comprehensible manner.

The following pages describe the in-car technologies which are of particular relevance from the point of view of IT security. They also outline the challenges that have to be overcome in order for artificial intelligence to be used safely and provide an overview of the regulatory framework and standards that currently govern cyber security inside vehicles.