Automotive cyber security: a real, present and growing challenge
Eric Chan, Ricardo connected and automated vehicles expert
These days, we take it as read that computer systems need to be protected against online threats, and the term cyber security is very much a part of everyday vocabulary; cyber crime is now criminal big business, costing the UK an estimated £30 billion annually. But the recent global ‘WannaCry’ worm provided yet another sobering reminder that we are still a very long way from achieving true security even in comparatively mature and mission-critical business IT systems. Perhaps of even greater concern is that the scope of the devices and systems that we need to protect is expanding at an increasing rate, as the ‘internet of things’ gives us all manner of internet-connected devices, from domestic central heating to white goods.
Connecting everything together, however, brings the risk that one compromised system will be used as a gateway to other. Confidential and potentially valuable data can thus be placed at risk on any other device connected to the same network, either full-time or on an occasional basis, from smartphones, home Wi-Fi routers, and network disk drives to PCs. Few consumers may be aware, but this list of internet-connected devices already includes our cars. Many new models are connected over the air for telematics and maintenance, for safety systems such as eCall, by consumers using insurance-based monitoring technology, and by the many smartphone apps available to consumers.
The most obvious motivation for the hacking of automotive systems is for theft of the vehicle itself, and there have already been instances of thieves taking control of key encryption through the use of pirated software and exploiting vulnerabilities in security systems. This could take the form of radio ‘amplification’ attacks to spoof keyless entry systems, for example. Beyond theft of the vehicle, other criminal motivations for hacking include the theft of personal information from payment systems, accessing data from onboard sensors such as cameras and lidar, taking control over vehicle functions perhaps in the manner of a ransomware attack, or using the vehicle as a compromised gateway into other connected systems.
The need to provide cyber security for the vehicles we own and operate is thus a crucial and growing imperative. Cyber security must in my view be considered during the design phase, as adding it on afterwards will never be as effective. This represents a major challenge for vehicle OEMs to transition their legacy vehicle architectures, ECUs, and development processes to take into account these new requirements. There are strong links between methods and approaches required in the development of functional safety and cyber security, offering opportunities to us for early progress. But there are important differences too. For instance, while the threats protected through functional safety remain constant through the life of the vehicle, cyber security is an ongoing war of attrition against constantly innovating criminals who will have access to new data and tools over time.
Fifty years ago, not least prompted by the efforts of campaigners such as Ralph Nader, the auto industry realized the need to place safety at the forefront of design, and in the 20 years since the advent of Euro NCAP, no one questions the need for safety any more. What we need now, in my view, is the same focus upon automotive cyber security – and this is one of the primary reasons for Ricardo’s participation in the 5*StarS programme (see Ricardo News p25) which will address the increased threat from cyber security with the proliferation of connected and automated road vehicles.
This view point featured in RQ Q2 2017 - click here to download the full publication.