Continental's top tech executive, Kurt Lehmann, says that making vehicles part of the Internet of Things is one of the biggest challenges that he faces. In an interview with Automotive News Europe Correspondent Olive Keogh he also shared his views on how cars could benefit from artificial intelligence.
What are the biggest future challenges for connectivity?
To make vehicles part of the Internet of Everything. The entire technology in and beyond the vehicle must be rethought to achieve holistic connectivity. It has to be seamless with the driver and with other vehicles and mobile devices, which gets complicated. The other big challenge is ensuring cybersecurity.
Can you explain what you mean by holistic connectivity?
Holistic connectivity is about more than just vehicle connectivity. While vehicle connectivity was once simply an add-on feature, it is now becoming a key technology for intelligent mobility of the future. This is why holistic connectivity concerns all aspects of a connected vehicle – from the electronics architecture in the vehicle, through the connection to the surroundings, to mobility services and their operation for the driver and user. For us, it means creating a seamless online connection and transforming the entire vehicle into an open system.
Does 5G have a role to play in this?
Yes. As 5G becomes more popular it gives us more options, although we will never be completely done. I don’t think we really have a view yet as to what it could look like in 10 or 15 years. We just know that in the next few years there is going to be a rising expectation by drivers that they will have the same level of seamless connectivity in their cars as they do in their living rooms or offices, with the same level of privacy and security.
What are the big challenges around artificial intelligence?
We know it can achieve better performance than traditional methods of deterministic software. However, you can't really explain that away because neural networks don't give up their secrets. That's a challenge because we’re dealing with safety-relevant applications. There is also the challenge of the computational power that enables AI. That goes back to architecture and moves us toward server-level or service-oriented architectures that are well known to the IT people but new for electronics within a vehicle. AI needs to be trusted and will take time to be proven, but I’m highly confident. It’s a question of time and gathering data. There are AI milestones and I think we’re on track to reach them, but we have to be careful about knee-jerk legislation in response to some of the events that have happened recently.