How BMW-, Mercedes-owned map specialist Here aims to stay a step ahead of Google

Right here Systems has productively fended off competitors from technological know-how giants hunting to transfer into large-definition mapping, which is a single of the critical factors for autonomous driving. CEO Edzard Overbeek lately informed Automotive Information Europe Correspondent Nick Gibbs how the locale information and technological know-how system supplier will preserve a move forward of Google.

Volvo and Google employed CES to announce a partnership on High definition maps. How do you battle the tech giants?

We do very very well. For instance, the Mercedes S-Course with Amount three autonomous functionality employs our large-definition maps. We just released with the BMW i7, which is presently Amount two Moreover and will go to Amount three. As much as I know, we are the major provider of High definition maps in business contracts. Google has a romance with Volvo, but we also have a romance with Volvo, not for High definition maps, but on other issues.

What are your strengths from another person like Google?

They will not have our abilities. For instance, in Europe we have 85 p.c of the marketplace for the new ISA [intelligent speed assist] regulation, which wants to know the pace restrict of the street you are driving on [to provide feedback to the driver if the posted speed limit is being exceeded]. The tech organizations will not have that. The 2nd is privateness. Corporations are commencing to notice that when you use Google Maps, you are fundamentally marketing your information to Google for them to create greater ads. We will not do that. The 3rd is that we have no concern of competitors. We are neutral Switzerland when it arrives to locale information. We are not undertaking just about anything else with it. If you do company with the Googles of the globe, then they may arrive into your business.

Exactly where are your most significant threats coming from in conditions of startups?

It can be a really suitable problem, but I would possibly have answered it in different ways 5 several years in the past when every person and their moms have been stepping into mapping. Currently, really handful of are remaining. Deepmap has been obtained [by chip provider Nvidia]. Civil Maps just bought obtained [by lidar company Luminar]. So, of all the startups close to mapping, I consider really handful of have survived.

They have been purchased but they nonetheless exist, so usually are not they a threat?

They bought absorbed into a stack, starting to be aspect of a vertically built-in method. That is not truly a mapping competitor for us. Individuals are commencing to notice the significance of locale information for all use instances, not just the autonomous car. Nonetheless, setting up this is really pricey and really sophisticated. We have found men and women test. And, indeed, it can be wonderful to set a lidar on a vehicle, travel Interstate 280 [a major north–south highway in the San Francisco Bay Area] and then say, “I have bought a large-definition map of I-280.” That is great, but now do that for 200 international locations close to the globe in close to authentic time with refresh costs of 24 hrs or considerably less in every single resolution variety.

Can you get information from consumer automobiles like Google will do from the new Volvo EX90 SUV?

Sure, we have 35 million automobiles supplying us sensor information that we make maps from. We extract the watch from cameras and make a electronic illustration. That is what we do at scale. In our new Unimap we have 500 million kilometers of roadways getting processed every single hour. So, it can be wonderful that Google has now entry to Volvo and I will not want to downplay that, but it can be a tiny little bit distinctive.

How are you downloading that info from automobiles?

We will not do the complete picture. It can be not the complete two terabytes of information for the reason that that would be uneconomical. Suppliers enable us entry from information feeds from the vehicle sensor cluster.

What do the producers get out of you extracting their information?

They get greater maps. They get locale-dependent solutions. We are sitting down down with them on a full array of new ordeals when it arrives to navigation. So, what is the subsequent technology navigation seem like? Is that only in the center console or are we heading to do OLED see-by shows and other shows in the vehicle?

Can you extract large-definition information from consumer automobiles?

Sure. You want lidar, which has millimeter precision. Radar and ultrasound are not large definition. They could be in the potential but not now. We have arguably the major lidar databases in the globe.

You introduced at CES a further way to supply items, Unimap. How does that get the job done?

In the previous we started off with a single pipeline — common definition maps. Then we create a further pipeline that supplied website traffic information. These are fundamentally automatic flows from a databases that steers into the machine — a vehicle, a cellphone, a drone, a robotic &#x2026 whichever. Then we included areas information as a new pipeline, then large-definition information, then 3D maps. What we have performed with Unimap is combine them all into a single pipeline.

What is the gain?

Now we align the information feed so they make rational perception to each individual other. 5 pipelines into a single. No person else has this in the business. For instance, the website traffic mild info was not joined to the street info. Now it is. So, the street segments will become considerably richer.

How does this advantage automakers?

They say this is what they want. With 5 distinctive feeds, it will become a spaghetti and with a really large charge to combine. Obtaining a single pipeline is much a lot more charge productive.

What takes place if another person just desires a single component of that pipeline?

You just extract, for instance, common definition maps or just street curvature or tunnels. You only pay back for that. It will become a considerably a lot more modular way to take in map information.

Is this all downloaded into the vehicle?

It is now but in the potential you can do that through the cloud.

What is This is automotive/non-automotive break up?

About 70 p.c automotive and 30 p.c non-automotive. The method is to increase the two. We have a very big share of automakers by now, but we are seeking to grow that. In non-automotive, our aim is on transportation and logistics, retail and e-commerce.

What is actually your geographic break up?

In all probability 40 p.c U.S., 40 p.c Europe, 20 p.c Asia. China is an crucial marketplace, but we will not function straight in China. We have a joint enterprise with NavInfo that serves the inbound [non-Chinese] automakers there.

Exactly where is the subsequent large progress in automotive?

We have a lion’s share in automotive. We have two new items, a single is street alerts, for instance, if the vehicle forward has damaged down, we can show that. We also have autonomous driving zones, which is fundamentally geofencing the map. If there is an setting in which you will not want the autonomous driving method to be energetic, you can make certain it can not be activated there. Or, if there are sure occasions in which you will not want the automatic driving method to be energetic, you can fundamentally shut it down.

What does a large-definition map indicate?

Maps are for human beings. A vehicle wants an instruction established it wants device language. That is in which we are transferring the business, towards a lot more of a locale, large-information business. A device just usually takes the feed of a lane or a street curvature or a bridge or whichever.

Some organizations have reported that palms-totally free driving could be attainable without having large definition. Are you viewing that craze?

More than the past pair of several years, sure organizations have reported sensors in the vehicle are very good plenty of, but I consider every person has stepped absent from that considered. The realization now is that possibly the most crucial sensor in the vehicle is the map, and that every thing feeds into that. A map seems to be close to the corners, ideal? So, in automatic driving, the vehicle will become mindful of the scenario and seems to be by structures and close to corners.

Is there just about anything automakers can consider from the large-definition map and change it into some thing appealing for the driver?

Sure, from a locale-dependent solutions level of watch. For occasion, consider about hunting at a setting up and downloading the background of that setting up on to your vehicle window. We are pondering alongside all those traces.

What does owning automakers as shareholders do for you? Does it defend you? How do they advantage?

It can be really strategic for them. They took the final decision to acquire the business for the reason that they imagine that the aggregation of information is critical for the vehicle business in typical, not just for the shareholders. It can be about producing certain that they are impartial from the large technological know-how organizations. I seem at them as shareholders. We have really distinctive discussions with them as shareholders than we do when we are speaking about business contracts with them, and we will not combine the two.

Do your shareholders often do company with your rivals?

Sure. There is no ensure that they give us the company just for the reason that they are shareholders. We want to execute. That is the way it can be established up. For instance, BMW has a offer with Mapbox, and Nvidia is a husband or wife with Mercedes on large-definition mapping. It can be my occupation to make certain that we are greater. It retains me on my toes.

The post How BMW-, Mercedes-owned map specialist Here aims to stay a step ahead of Google appeared first on Cars News Magazine.

The Auto Xone