AI Online


Security holds the key to opening an exciting new world of vehicle connectivity that will herald the introduction of autonomous cars. The more vehicles become connected however, the more entry points you open up for potential hackers. 

Automotive Industries (AI) asked Andreas Mai, Director, Smart Connected Vehicles at Cisco why vehicles are at such risk.

Mai: The problem is that vehicles were not designed for security. In particular, the CAN Network is very badly equipped to provide sufficient levels of security. What is needed to manage the risks is a joint initiative where the industry gets together to define the severity of certain threat factors and threats, and finds a standard that the industry can use to fend off cyber-attacks.

AI Tell us a little about Cisco M2M applications.

Mai: If you think about the world that automakers live in right now, the electronic content by value is reaching 30% to 40% of the vehicle. This means there is a lot of software involved in modern vehicles. This software traditionally has glitches that need to fixed/ updated. The ability to identify and address these issues before they cause any harm will save the industry millions of dollars. The second challenge is to updates these connected vehicles remotely. The important aspect to consider here is security. You don’t want anyone to intercept and interfere with the updates to the vehicles.

AI: What new technologies are being introduced in advanced driver assistance systems (ADAS)?

Mai: Cisco has invested in Cohda Wireless, which provides the most advanced dedicated in short range communication units with ultra low latency communication for V2V and V2I.
This potentially life-saving sensor gives the driver one or two seconds warning before they would see an obstacle on the road. Another technology that we see being used in the context of ADAS and the autonomous vehicle in particular is greater use of IP (Internet Protocol) in vehicle networks. To enable this you need to have a lot of sensors onboard the vehicle and you need to merge the impulses from these sensors in real-time. The existing in-vehicle networks have severe limitations, which is where we see the role of IP.

The other reason you need higher performance networks is that both ADAS and autonomous vehicles are utilizing imaging from video feeds that provide a 360-degree view around the vehicle. This information is analyzed in real time and fed through to the vehicle control center. The last element is security: the CAN network has design limitations which affect the keys – such as the 256 key – which are used to manage security. IP provides significantly more room to embed security keys.

AI: What are the benefits Cisco cloud services offers to OEMs, dealerships, and the end user?

Mai: In a world where the vehicles are connected and you provide updates over the air, you may want to use the dealer Wi-Fi networks for large files. Cisco has developed an end-to-end security architecture for protecting devices, and in particular mobile devices such as cars which are driving around in open, uncontrolled environments. Cisco is also adding value through our “Intercloud”. One needs to stitch to together the individual clouds of the OEMs, Tier 1s and dealerships in order to allow them to collaborate efficiently.

AI: How do you see the future of vehicle communications technology?

Mai: In the past, the industry has looked for a single technology – do I put 2G, 3G, 4G or Wi-Fi in my vehicle? This has evolved into developing a platform which connects to all them, even DSRC and satellites. The nature of a car is that they move across networks, so you need a solution that delivers ubiquitous high performing connectivity to vehicles. I see the technology evolving whereby it smartly selects the best uplink, depending on the application