When we looked at the problem of how to secure a connected car, we had three goals in mind in building our solution. The first goal was that the solution had to be able to protect against remote attacks that are generated over the internet. The second goal was that we wanted to protect both the single vehicle as well as the entire vehicle fleet, expanding across multiple make models at the same time. And the third and final goal was that we wanted to be able to protect vehicles that are on the road today, not vehicles that are gonna leave the factory in two to five years’ time. So how do we go about solving this problem?
What we realized was that the only way to address all three of these requirements is by using the cloud. In the cloud, what we’re doing is we’re leveraging data that is already being collected by the OEMs and connected vehicles. The cloud also allows us to rapidly upgrade our software and make sure that we’re always two steps ahead of the hackers without having to rely on the upgrade cycles of the existing vehicles. Once installed, our platform collects data from all three of these sources. We start with data coming off of the TCU from the connected vehicles. We add typically the telematics server that’s hosted in the automotive cloud. And finally, we introduce the data coming off of the mobile application server.
This is the point where the Upstream cloud platform actually goes to work. What we do with all this data is we analyze it using advanced machine learning and big data analytics capabilities, and we model the entire connected car service. The platform understands what’s the normal behavior of both the application servers, mobile, telematics, any additional service, as well as the behavior of any given car within the connected car service. Once we understand what the normal behavior is, we’re now able to detect any violations, anything that’s outside of the norm. We define what normal is starting from the protocols that the automotive cloud uses, all the way up to the behavior of the app servers and the vehicles themselves.
Once we define what the norm is, we’re able to detect what is outside of the norm and use that information to create incidents. The incidents are then being consumed by the security operations center and the various security analysts that operate that. At the end of the day, our product is being used by the SOC team, the various security analysts. It gives them a new visibility that they never had before and the ability to detect incidents in real time and perform triage and root cause analysis, and actually get to the bottom of things that are happening in the connected car service.