February 17, 2040. My autonomous taxi drops me off at work.
Today, I have an important meeting with a client. I’ve had my cup of tea and I am heading to the meeting point to take my autonomous taxi. I get in this autonomous vehicle programmed to drop me off in the city center.
I am serene and have full confidence in this intelligent car. I know it will get me safely and faster to my destination using the most optimized route. It is not only equipped with several sensors but during the course, it is constantly exchanging data with other vehicles and road infrastructure to keep me safe. I will be reaching my destination soon.
While approaching an intersection, I see that ALL the traffic lights are green, but my autonomous taxi suddenly stops. The truck coming from the right also stops. What’s going on? Why did the vehicles decide to stop? Is it a malfunction, a cyberattack? Luckily, faced with this abnormal situation, my autonomous taxi decided to halt, even though the light was green.
Misbehavior detection, a prerequisite for the deployment of autonomous vehicles
Today’s vehicles are increasingly equipped with technologies that allow them to interact with their environment. These intelligent vehicles exchange information with each other and road infrastructure to improve road safety, traffic efficiency and users’ comfort.
Communications between vehicles and with the road infrastructure must be secured for the data exchange to be reliable.
A security solution called Public Key Infrastructure (PKI) already makes it possible to protect these communications by providing digital identities to each entity of the system. Digital trust, guaranteed by the PKI, ensures the integrity and authenticity of data exchange through cryptographic mechanisms.
But it’s not enough. As autonomous vehicles will make decisions based on the data exchange, this data must be reliable. Let’s imagine that a hacker takes control of a traffic light and changes its color; he will be able to send false information over the network. An autonomous vehicle could also, in case of technical failure, send erroneous messages.
To enable misbehavior detection, a solution must be deployed within each vehicle and road infrastructure, and a central system must be set up.
The future autonomous taxi will be equipped to detect misbehavior. It will receive a message from the traffic light, and it will check it. With the help of its sensors and messages sent vehicles in the vicinity, like the truck, it will be able to detect inconsistency of the data received. It will then make the right decision and determine the suitable local reaction. The autonomous taxi will send a report to the central entity. The process of misbehavior detection will also be triggered by the truck that similarly witnessed an abnormal situation. Cooperation between the actors of the system will play a key role.
A central entity will be responsible to receive misbehavior reports and to investigate. Its purpose will be to determine whether the traffic light has been compromised or not. If so, a global reaction will be triggered. The possible responses can range from a simple notification to a control center (to request a maintenance intervention ) to the exclusion of the entity from the trusted system. Messages sent by this entity would then no longer be considered by the others.
Artificial intelligence takes cybersecurity to the next level
The misbehavior detection solution will benefit from using artificial intelligence, such as machine learning.
The autonomous vehiclewill have to process a large amount of data in a short time to detect an anomaly. The central system will have to gather a very large number of reports, extract suitable data that will allow to predict the type of misbehavior and establish a more appropriate reaction.
Artificial intelligence enhances digital trust and will be one of the sine qua non conditions for the deployment
of autonomous vehicles on our roads.
Discover how Atos answers the challenges of Intelligent Transport Systems (ITS) with its IDnomic C-ITS PKI solution to generate and distribute secured digital identities to C-ITS stations.
Bring trust to Intelligent Transportation Systems with a cybersecurity and standardization approach.
Interested in next publications?
Register to our newsletter and receive a notification when there are new articles.