HERE Automotive on strategies to support autonomous driving - Just Auto
Join Our Newsletter - Get important industry news and analysis sent to your inbox – sign up to our e-Newsletter here

HERE Automotive on strategies to support autonomous driving

01 Feb 2016 (Last Updated July 2nd, 2021 20:30)

At the 2016 CES, Cat Dow talked with Floris Van Der Klashorst, vice-president of HERE Automotive. HERE's a creator of connected navigation solutions and since December 2015 has been owned by an alliance of the German automakers Audi, BMW and Daimler.

At the 2016 CES, Cat Dow talked with Floris van de Klashorst, vice-president of HERE Automotive. HERE’s a creator of connected navigation solutions and since December 2015 has been owned by an alliance of the German automakers Audi, BMW and Daimler. 

just-auto: What’s changed since the acquisition by the alliance?

Floris van de Klashorst: It was more of the perspective of why they bought us, acquired us, which is important. If you look at what we stand for, as a company, and what we try to establish, is really being a pure location player, posting the content in the data analytics, in the software platform, and the applications in the cars. 

If you look at HERE as a company, we do business at multiple layers with a lot of customers. There’s more than traditional business, which is the likes of content for navigation, but if you look towards other horizons, such as connected cars and autonomous driving, there are different requirements and different needs. 

That’s what we recognised around four or five years ago, when Navteq and Nokia services joined together into a new company, and we rebranded it as HERE, where we started to focus on a pure location platform, with the software on a cloud-based platform. 

That adds a lot of intelligence and value on top of the content, based on location data, and the first instance for the change in strategy or extended strategies, building a software stack and application stack on top of the content, have been in HERE Auto or in embedded navigation systems.

This predicts where you’re going, is personalised and understands what vehicle sensor data gives, in terms of fuel and other factors, which is a smarter system than the traditional ones. It’s also working towards another long-term horizon, which is autonomous driving, and what you see is that the vision, the strategy and what we have developed to support that, is exactly the reason why the consortium bought us. 

So the implication of the acquisition is that we now have immediate access to more high-quality data from premium cars, which have a lot of sensors, so it will accelerate our strategy and our goal out of our products, which is very good. But that’s one side, together with our new owners, who have an interest to remain independent from the likes of Google and others. 

So they are investing to help us with that, but on the other side of the spectrum, it’s a collaborative system, when it comes to autonomous vehicles. There’s not a single car manufacturer in the world that has enough cars to maintain a system of real-time data, which has observations of incidents and accidents, so it’s where you need all carmakers. 

The way that they have set up the governance to keep us independent and open, as well as the fact that they are getting other parties to join the consortium, is very important for our strategy to build a system of standardization for the entire industry. So while they have to capitalize and invest to work with us, they have also come to recognize that it is better for them to keep us open, and that is the strategy that we are executing. 

j-a: What have you brought to CES 2016?

FVDK: Today at CES, we launched our HD Live Map. This is the highly-precise map for autonomous vehicles, so there is no other product that matches it, in terms of accuracy and it’s really designed for the automotive industry. 

j-a: And how does HD Live Maps fit into your current strategy, as well as medium to long-term?

FVDK: We launched the first version of it is which covers all the connected data use cases, like speed curvatures and profiles, etc, for Western Europe and North America. We are also continuing to work on selected areas for OEMs that really want to test for the next year or so. 

In 2016 to 2017, we will deploy to more countries, so it’s more now for the HD Live Map to supportive the automotive industry. It’s not a product for consumers, so we are focusing more on the planning for the different content partners that we have.  With regards to consumers we have a new product coming out reference mobility needs. 

j-a: Tell me more about them.

FVDK: It will become more complex for us in terms of cities and options, taking your own car, car sharing. For example, the ability to select public transport to go to an ATM to withdraw some money at a public railway station near me, that kind of search when it comes to planning, which we called Maplinks, it uses a playful combination of different types of information to support your immediate mobility needs. That’s really our consumer strategy.

j-a: And how do you intend to take the battle to Google, which seems to all intents and purposes, focussed on similar deliverables?

FVDK: We’re not trying to mimic Google, or be better than them. We’re trying to be more relevant to consumers with the application. If you’ve had disappointments in data accuracy with the app, in a consumer application you can always update, but if you have a look at the HD Live Map, it is an immediate, self repairing map, and that is where we’re going with our strategy. 

If you look at the traditional update mechanism, it is always slow. What Google doesn’t have to do, is that they don’t have to validate. It because they have no liability, no claims, as well as being a free consumer product, so it is not always accurate. So the data that we give our automotive clients, that we also use through our consumer app, has already been validated. 

j-a: Has that validation process caused issues previously?

FVDK: In some cases, we may be slower with data than Google, but we do weekly updates now with all the data, because the past compilation processes took a lot longer. But we do not allow consumers to put in changes that are not validated, because our data also goes into the car. 

It is a different business model, where we have different requirements for the OEMs to adhere to, and sometimes that leads to a difference in the freshness of consumer data that you experience through the apps. 

j-a: So in terms of mapping from an autonomous driver perspective, are you on track with your objectives and goals? 

FVDK: Yes, we are, having worked with more than ten OEMs on concepts and experiences over the last two years, and so should the HD Live Map be, in terms of definition and specification. We have published our findings on that, based on the feedback of the OEMs, as well as launching our data ingestion platform publicly a couple of months ago.

So it is a platform for the OEMs to contribute their data, and now we are rolling out the commercial availability of HD Live Map with all the collected data attributes for first use cases. 

j-a: So do you think 2020 is an ambitious target for some of the manufacturers?

FVDK: Some of the collected data areas will be available earlier than that, I think it’s too ambitious for becoming fully autonomous, there will be highly used autonomous cases by 2020 by multiple OEMs.

j-a: Are you able to give me any insight into regulatory bodies that you have been working with, where you can see that there is going to be an upcoming change? Some of the technology that is going to be used is already here. With the mapping system already in place through yourselves, so all we are waiting for is a change in regulations, no?

FVDK: I think it’s not just a case of change in regulations, but I think it’s also experience inside the car, which we are working on. Maybe the technology is there, but one of the important crucial elements that we are working on is consumer acceptance. It is still a person stepping into  a car that needs to trust the vehicle to do the right things.

The HD Live Map has three value propositions. One is this highly accurate map, but that is only really for the correct positioning of the vehicle in longitude and latitude, the number of lanes on the road, the exact location of exit ramps, traffic lights, etc.

What’s happening on top of the roads is beyond the reach of the on-board sensors, beyond the cameras and the radars around the corner, in distances exceeding two miles beyond, sensor data can be captured via other cars and then broadcast through the system itself for those cars that have not yet reached that location. 

This means the cars can change lanes and reduce speed earlier, which is called the life activity data, on top of the HD Live Map. So that is the second value proposition.

The third proposition is the profiles, such as speed and driver profiles, showing how people actually take corners. Say, for instance, the legal speed limit is 70 km/h, but people never drive faster than 50. It also makes the car feel more natural in how it drives itself. It’s not only the capability of the car to drive itself that we support, the capability to take better informed decisions earlier, so the behaviour and decision making of the car  installs both confidence and trust in the people that get into it. 

Regarding the legislation in the countries, you can see in the marketplace that other countries want to get there first with the licences being issued, wanting to lead when it comes to autonomous driving. The level of alignment between what the technology can offer, in terms of better usage of the capacity of the network, reduction in emissions and less casualties, shows that the alignment of those capabilities with the social agenda of governments is tremendous.

j-a: How do you see it developing?

FVDK: I see more support and more forward leaning, especially with the ITS conference in Bordeaux that took place in the Summer of last year. I was on the panel along with road authorities from Europe and the USA, who were all leaning forward to make it happen. They all saw the need for standardisation on continents, so you don’t have different systems competing with each other, and we are still having those very lively conversations with them at present.

We are seeing a lot of support, which has good momentum to make sure that liabilities and other factors are taken care of, carmakers showing their interests saying that they do want this.

j-a: So this means that the second layer the system relies on the cloud?

FVDK: Yes, in association with all the car manufacturers. The infrastructure for the delivery of cloud services is again somewhat limited in capacity to what we would ideally like it to be for the full-on specification of these systems, including dead spots for 4G signals.

j-a: What are the challenges have you faced in bringing this kind of mapping product together, to demonstrate it here?

FVDK: Most of the information that we have is not time critical, as the car should always be able to stop safely or avoid a collision without connectivity. This is because the connectivity could break or you could have a signal dead spot. 

So the HD Live Map is not the system that prevents a collision, but it’s a sensor that allows better and more informed decision-making for the car. It allows for a much more intelligently based manoeuvre, but if it’s not there, the car can still drive by themselves. 

This has been tested by Google and you can also see what the limitations of those systems are. That’s what HD Live Map brings, which completes the circle of information that a car needs in order to drive in a comfortable and secure way.

The safety of not hearing something is always based on internal duplicate sensors, which are contained within the cars themselves. 

j-a: Finally, what’s the next step?

I think the next step is the car manufacturers submitting the data to this platform, where we create a large pool of information that everyone benefits from. The three consortium owners have taken the first step and we expect others to follow.