A review of the relationship between cyber-physical systems, autonomous vehicles and their trustworthiness

A concentration of development in cyber-physical systems containing advanced sensors, sub-systems and machine-learning algorithms over the past decade is equipping unmanned aerial and road vehicles with autonomous decision-making capabilities. The level of autonomy depends upon the make-up and degree of sensor sophistication and the vehicle’s operational application, which can range from extremely high cost military-based combat vehicles, to commercially available civilian motor vehicles and very low-cost sub $100 hobby drones. As a result, the risk of autonomous vehicles being compromised and used as an improvised threat has become more serious and therefore trust-based methods to mitigate them are needed. This paper reviews the relationship between cyber-physical systems, automated and autonomous vehicles and their trustworthiness. It examines the significant role cyber-physical systems play in building sub-system, system, and system-of-system layers found in contemporary vehicles as well as the level of human involvement in decision control loops, and the trustworthiness challenges this creates across levels of vehicle automation and autonomy. Further, the paper reviews material on methods to authenticate a vehicle’s “world-view” and the veracity of information it may communicate and share with other vehicles and infrastructure. It finds that to be trustworthy, automated and autonomous vehicles will need to implement data verification frameworks and methods. It recommends that future research be conducted into the effectiveness of implementing a pervasive trustworthiness hub where information flowing between heterogeneous autonomous vehicles and infrastructure within their surrounding environment can be authenticated.