Volker Hartmann (44) is the new legal boss at the Munich startup Fernride Fernride
Hardly anyone is as familiar with the laws for autonomous driving in Germany as he is: Volker Hartmann. He has more than 15 years of experience in the automotive and mobility sector and in the field of autonomous driving, including at Audi, Daimler and in the VW Group at the software arm Cariad. Most recently, he worked for the telefahr startup Vay as Global VP of Legal and Governmental Affairs. Now he has switched to rival company Fernride, as General Council.
Vay and Fernride are the only two startups in Germany dedicated to so-called tele-driving. This means that the vehicles are controlled remotely, like in a computer game. The teledriver sits in a control center and steers the vehicle with the help of numerous cameras and sensors.
So the technology is similar, but the business models of Vay and Fernride are completely different: Vay wants to set up a partly remote-controlled taxi service, Fernride is developing software for trucks that only move on company premises. Vays Service is to become an end customer business, Fernride specializes in B2B.
The 44-year-old Hartmann does not want to talk about his previous employer, but he does want to talk about the major challenges in the area of tele and autonomous driving – from moonshot scenarios to warning triangles to the “bottom meter”.
read too
Fernride gets $31 million – and takes a different approach than competitor Vay
Volker, your last employers have different business models. What excites you about the B2B and logistics area in which Fernride is active?
This is where the need and the overall economic benefit are greatest. All of our prosperity depends on the logistics chains functioning. Hardly anyone wants to work as a truck driver today, and that will be even more the case in the future.
That’s why it’s so promising to automate the field.
Yes, this is how we can counteract the shortage of skilled workers, among other things, exactly. At the same time, I have an area on a closed company site that I can control relatively well. That doesn’t mean it’s completely trivial and we can do whatever we want. That’s why I’m here now. But it’s easier to control than following a moonshot scenario.
What do you mean by moonshot scenario?
By that I mean the approach in autonomous driving, of consciously wanting to solve the most complex of the most complex things right from the start. Companies like Argo, Waymo, Zoox or Cruise, but also the big car manufacturers have often followed this approach. Of course, it is good to set yourself an ambitious goal.
But?
I find this approach difficult, if not impossible, from a security and regulatory perspective. And it has proven unrealistic over the last few years.
Do you have an example of this?
You can see that with automated vehicles, for example in San Francisco. They just stop and do things that no one expects. Because when in doubt – and rightly so from a legal point of view – they are so defensively programmed that they do not want to break the law and simply stand still when they are unsure. But that in turn is unsatisfactory and can sometimes create dangerous traffic situations.
So it would be better to use a human as a teledriver so that the vehicles don’t stop?
The biggest challenge for the logistics operation is: How do I manage to keep the trucks moving and thus operations on the company premises? If an automated truck has to bring something from A to B and it just stands still, this can trigger a huge chain of complications. When in doubt, it is much better to let the teleoperator take over, even if it is economically more expensive to keep bringing people in with you. But until I really master this use case perfectly, I’d rather do that than break the chain.
In the start-up scene in particular, the legislative processes are not fast enough for many, especially for new technologies such as autonomous cars. neither do you?
I’m a big fan of there being good, appropriate regulation for new technologies. Many say that makes everything more complicated. But if regulation is done well, then it can be more conducive to the advancement of technology and serve as a safeguard. After all, an engineer wants to know where he is and not have to worry about ending up in prison for his latest development.
So the current laws on autonomous driving in this country are not as retrograde as is often claimed?
The current laws in Germany and in Europe with regard to autonomous driving are definitely a step in the right direction, even if I’m not one hundred percent satisfied with all the details. There are also gaps, for example in the law of conduct – what does the robo-taxi do when it has to set up a warning triangle? Or when it comes to direct, teleoperated driving.
So there is no legal basis for tele-driving?
At least not explicitly – and here we are again with my philosophy that good regulation is something positive.
read too
Embellished safety data: did the telefahr startup Vay get an exemption?
Experts see only a kind of interim solution in tele-driving until there are fully autonomous cars in Germany. When will that be?
From my eight years of experience in advising various technology projects in autonomous driving: Level 5, i.e. autonomous driving of a vehicle in all application areas and without a driver present, will remain a wishful thinking for a very long time.
Very long means: the next ten years or twenty years?
That’s what I’m expecting. In places where you can really set up your own infrastructure for autonomous driving, for example in China or in countries like the United Arab Emirates or Singapore, things might go faster. There, practically laboratory conditions prevail in demarcated areas.
That doesn’t exist in this country?
In Europe or in North America, we cannot make any significant changes to the infrastructure for various reasons. In Berlin, for example, we couldn’t simply flatten the city autobahns and only let autonomous trucks drive. That is why we have to develop hybrid mobility concepts and need approaches such as tele-driving in order to be able to deal with mixed traffic.
read too
Not Tesla: Trucks are the real drivers of autonomous driving
What’s so difficult about that?
The rule of thumb is: the more mixed traffic there is, the more complex it becomes – i.e. autonomous and non-autonomous vehicles, cyclists and different types of pedestrians, such as children, disabled people, old people, maybe drunk people. Can the system recognize from the body inclination of the respective person whether someone is just tired or perhaps drunk? Is someone stumbling or is it likely that this child will start running? All this is an extreme effort that has to be made in order to reliably map such complexities. That’s why the idea of teleoperation as a solution for such exceptional cases, to say: Hey, I can simply fall back on the human perception and judgment of the teledrivers.
But telephonic driving is not perfect either. What are the biggest challenges?
The challenge there is to design the technology so well that, despite the latencies that may occur in the video transmission or the 2D display on the screens, the teledrivers can classify the events just as you or I would if we were actually a would chug along the busy play street. The Fernride founder Hendrik Kramer once called it the “Popo-Meter”.
So the haptic driving experience that you have when you actually sit in a vehicle?
Exactly, that’s actually a great term. But it must also be clear: Depending on the application, teleoperation can be almost as complex and time-consuming as automated driving. What happens, for example, if the connection to the tele-driver breaks down? Does the vehicle then simply brake hard while driving?
Then you would still need an autonomous system that can take over if necessary.
Exactly. But if, for example, I’m in the middle of an intersection and then slam on the brakes, that might not be such a good idea either. This also does not comply with our current traffic laws. Actually, you would need your own little robot as a backup, which recognizes itself: What is actually happening here right now? Is the error so bad that I really have to slam on the brakes? And is someone driving in front of me, behind me, next to me? And then react accordingly.
In what other areas do you think we will see automated driving, apart from logistics?
Military, agriculture and logistics – that’s the order in which I think automation will find the greatest application, and there’s already a lot going on there. Of course, no one is a fan of military use cases, but times have changed here, too. These are the domains where automation is being pushed the hardest and where it makes the most sense for society as a whole.
“>
External content not available
Your privacy settings prevent the loading and display of all external content (e.g. graphics or tables) and social networks (e.g. Youtube, Twitter, Facebook, Instagram etc.). To display, please activate the settings for social networks and external content in the privacy settings .
Change privacy settings