The connected car market is growing very rapidly. Leading market studies describe a growth rate of about 45% by 2021, which is reminiscent of the growth rate of the Internet market in its early steps. This growth predicts that all new vehicles sold will be connected by 2022. Actually, this is not surprising given that the various stakeholders, car manufacturers in particular, have identified the tremendous potential in this technological development, and no less important – the danger to the very existence of their business if they miss the boat. Therefore, everyone is rushing forward to meet their aggressive goals.
The center of attention in the connected automotive world are the autonomous driving applications which, without doubt, present an amazing picture. Essentially however, these applications are based on a network of computer-linked driving systems that can be described as a "gun in the first-act”, to borrow a metaphor from the theater. The next steps are to examine how these systems and applications stand up to security challenges.
Several questions arise at a time when technological advancement is critical to the success of companies' business: What is the meaning of connecting vehicles, in terms of exposure of internal systems to the outside world? Is it possible that connecting these driving systems with new applications will create safety exposure? When all the new computer-linked driving systems are put into the car, does anyone really know what is happening "under the hood"? Could a driver be accused of improper vehicle activity when actually the root cause is an illegal activity of the internal vehicle systems? What is the reason for a particular vehicle activity?
The rule to remember is that in every developing industry, the pace of technological development is faster than the rate of development of threats, and both are higher than the rate of development of defense technologies. Thus, a technological arms race emerges to cover the gap, as shown in Figure 1. The technological race creates an exposed operating area in the systems, sometimes dubbed "Death Valley".
We can take as an example the evolution of information technology and networks industries over the past decades. The technological developments led to countless cases in which the organizations employing it were exposed. After each event, an improvement was made in the security technologies to prevent the recurrence of the event. This virtual world was relatively forgiving to security failures, because in many cases the damage was not tangible, the systems returned back to service, and the information was restored. Annoying, but containable.
In the automotive world, the pace of technological development seems to be higher. More critically, there can be no forbearance or forgiveness for any event in a field that combines the cyber and the physical worlds: every case of automotive malfunction resonates tremendously because of the safety concerns, and the huge business damages caused to the manufacturers. An example is the extensive coverage given to the various Tesla mishaps due to the connected computer system failure and the security issues detected.
Because the connected vehicle industry is in its early stages of development, most of the possible threats are still unknown, so conventional cyber security technologies, based on detection of known threats, have limited capabilities. Without a complete solution, the problem remains, and the car companies find it difficult to move forward because of the difficult questions raised by safety bodies and the fear of the negative consequences that a malfunction might cause.
Artificial intelligence technology is needed for detection of unknown threats. Giant companies such as IBM and Amazon are recruiting employees who have knowledge in this field, but they are concentrating on solutions in the enterprise and industrial domains, and these do not provide a suitable answer to the needs of the rapidly developing connected vehicle world. Thus, this challenging task, which requires pioneering technological expertise and capabilities, is lead today by start-up companies in the field of connected vehicles security.
The Israeli cyber security market map, recently published jointly by Verizon and Deutsche Telekom, reveals that security solutions for protection of vehicles against unknown threats are already here. Beyond cyber companies providing cyber security solutions to known threats, some of which have already been swallowed up by global vehicle suppliers, there are companies such as SafeRide Technologies Ltd., which offers breakthrough anomaly-uncovering systems that can protect against both known threats and unknown threats.
SafeRide is led by some of the world's notable cyber security and artificial intelligence veterans, who have already produced security and analytics solutions for global technology giants, including Cisco and Microsoft, as well as car manufacturers such as General Motors and Daimler.
The technology that SafeRide offers includes software that can be installed on the connected vehicle systems and protects these systems against attacks, as well as protecting the vehicle from those systems if they go out of control through a hostile takeover. In addition to preventing known cyber-attacks, the system can study “normal” vehicle activity, create a behavioral baseline profile, identify unusual or ineffective vehicle operations, and warn of deviations occurring in a security, safety or efficiency context. This capability is provided by an artificial intelligence-based software subsystem that analyzes the vehicle activity and resides inside the vehicle or on the cloud, depending on the available vehicle compute capacity.
To develop such a system, the company’s data science team developed advanced artificial intelligence models as part of an anomaly-uncovering system they called vXRay (short for vehicle X ray), and trained vXRay to recognize “normal” vehicle activity. This process was carried out by analyzing extensive amounts of operational data for various types of vehicles, in cooperation with leading transportation providers. Data collection and vXRay calibration proved effective after the development team that built vXRay was amazed to discover that beyond detection and threat prevention capabilities, vXRay was able to predict future vehicle activity even before the activity occurred.
Advanced technologies that provide an answer to unknown threats, such as those offered by SafeRide, may for the first time herald a balance of power in the automotive technological arms race. The interest of car manufacturers and service providers seeking to adapt their technological base to the challenges introduced by the emerging connected vehicles market, reflects not only the need for a cyber security solution for connected vehicles’ exposure, but also the need to lead the connected vehicle revolution.