Exploring Local
Mike Dobson of TeleMapics on Local Search and All Things Geospatial

3D Road Geometry Requirements for Automotive Safety and Energy Management Applications – Part II

December 1st, 2009 by MDob

Ouch – this is really a long one, but on a very interesting topic. Get two cups of coffee and have a go!

Last time, I began reporting about the session I chaired on ADAS and Vehicle Energy Management at last months TeleMapics Update Conference in Europe. Just to refresh your memory:

The panel I chaired consisted of two representatives from companies interested in supplying high accuracy 3D road geometry data (Intermap and TeleAtlas), one automotive OEM (Ford), and one Tier 1 supplier (Continental) to the automotive industry. I thought that this mix of perspectives would provide a variety of opinions during the 40 minutes available for discussion. The panel assembled to discuss the topic included myself as moderator, Eric DesRoche, Senior VP of Automotive (Intermap), Christian Ress, Technical Expert, Telematics & Navigation (Ford), Christopher Wilson, Director of Strategic Research (TeleAtlas), Wieger Engbrenghof, Project Manager TeleMatics (Continental). In addition to his role at Ford. Christian Ress is the new chairman of the ADASIS Forum. Finally, NAVTEQ was invited to participate, but declined. Several members of the NAVTEQ team were in the audience for the presentation, but did not ask questions during the audience participation section of the discussion.

I prepared and shared 5 questions with the panel members well in advance of the meetings. Although I tweaked the questions a little the night before the discussion to make them less academic and more interesting, I did not change the nature of the queries.

So let’s see what was asked and answered. (Note: the following report is based on my notes from the meeting and these notes are directional, not comprehensive. If any of the participants happen to read this blog and want to add anything, please feel free to make a comment.)

Question 1 (a softball to start)
“What is your perspective on the advantages that high accuracy 3D Map Data might bring to Automotive Safety and Energy Management applications?
For Ford and Continental – if you have a positive outlook on the use of 3D Road Data, when might we see Automotive Safety or Energy management applications, supported by 3D Road Databases, available in vehicles?”

All panelists felt that high accuracy 3D Map Data was a critical component of future automotive safety and energy management applications. It was noted that these enhanced map data could be used to supplement and extend existing sensors used in safety systems and that there were other applications that would depend on these data as a primary input.

The speakers from the automotive industry indicated that the question was not whether these products would be brought to market, but when. As you know, getting a feature into a car takes a three year cycle and safety applications require more complex testing than some additions. However, it was clear that there was intense interest in testing these data for use safety and energy management applications. In other words, no surprises here!

Question 2 (A fastball)
“Does a 3D roads database need to be comprehensive across all of the functional classes of roads from Autobahn to minor local roads?

I ask this question because the lack of map coverage impeded the adoption of navigation systems and I am wondering if the same will be true in the development of 3D Road Databases. – especially since the category for minor local roads represents approximately 80% of the total road network and the majority of the database compilation effort.

For Intermap and TeleAtlas – what is your company’s perspective on this issue and what are you doing to address this issue?

For Continental and Ford – Is it possible to produce a viable product restricted to major roads carrying high volumes of traffic? Or perhaps more directly – at what level is the map coverage of transportation network adequate to meet your goals for safety and fuel efficiency applications?”

There appeared to be a difference of opinion between Intermap and TeleAtlas about the need for comprehensive road coverage. Eric DesRoche, Intermap, indicated his company’s position was the all classes of roads needed to be included in a 3D Road Geometry database, since safety and fuel efficiency applications should work on all classes of roads, not just on major roads. He added that Intermap was collecting 3D road data for all road classes across the geographic areas they covered. Conversely, Chris Wilson, TeleAtlas, indicated his belief that the initial 3D Geometry databases might be focused on major roads, with roads of less carrying capacity added at a later date.

Christian Ress from Ford was adamant that all classes of roads needed to be included in a database of 3D Road Geometry to be used for automotive safety and energy management applications. For instance, he indicated that, according to his research, Function Class 5 roads (minor roads) accounted for approximately eighty percent of Europe’s drivable roads. In addition, he noted that these roads are often narrow, hilly, curvy and unsafe. At the end, he seemed to twist the dagger, a little, indicating that “Roads are there for a reason.”

Wiegar Engbrenghof, Continental, concurred with his colleague from Ford, but noted that today no one map database vendor has all of the data (both 3D and other road attribute data) needed to support these applications and urged to vendors to cooperate and work together to support the automotive industry. My belief is that this plea reflected the current state of the data available to the automobile vendors. It is likely that they are conflating data from the vendors in hopes of creating comprehensive data sets needed for these advanced applications and finding that it is a very difficult task. Instead, they feel that the vendors should cooperate to provide the industry with the comprehensive data needed. This is a topic that I will discuss later in this blog.

As the discussion continued, I was surprised to hear the participants from the automotive industry indicate that while they preferred full coverage of all road classes in the 3D Road Geometry database, they fully expected that accuracy/quality would likely not be sufficient at the start, but would improve in due time. I had not expected a concession to quality at the introductory stages, but perhaps this reflects the automotive industry’s understanding that supplying the uniformly high level of data accuracy required is going to be a challenge for the mapping industry.

Near the end of the discussion, Eric DesRoche told that audience that Intermap had already released 1.4 million kilometers of 3D geometry data that included data for all roads throughout Germany. As panel moderator you can see what interests the audience and when Eric announced this tidbit, half the audience grabbed their pens and wrote notes about it.

Question 3 (a fastball)
“There appears to be various ways that 3D Map data might be made available to safety and fuel efficiency applications.

One scenario has 3D Map Data co-existing with navigation system and provided to safety and fuel efficiency application when required. A second approach removes the application dependency on a navigation system by storing the 3D Map data independently and allowing its access directly by the application through the CAN bus.

Do you see an industry preference for 3D Road Data being integrated with Navigation Data?”

If you read my 3-part article (starting here) on NAVTEQ, ADAS and MPE, you know why I asked this question. In October 2008, NAVTEQ introduced a strategy for a Map and Positioning Engine (MPE) that would provide a path for cars not equipped with navigation systems to use high accuracy map data for ADAS and other applications. It was my opinion, at the time, that NAVTEQ was looking for a lower cost path to support safety applications than one that required a navigation system as a base before safety applications could be added. Since I wrote my articles on MPE, I have heard very little about it in the market, although NAVTEQ is clearly working on populating their database, even though they have restricted it to the largest roads (function classes).

As might be expected, both Intermap and TeleAtlas indicated that they would support either stand-alone or navigation system supported applications. However, it was clear that representatives from Ford and Continental simply did not have a preference either for integrated or stand-alone databases to support their safety and energy management applications. In fact, they were more interested in testing comprehensive 3D Road Geometry data and once again urged the vendors to begin working together to provide the required data – as noted, a theme I will return to towards the end of this blog.

Question 4 (a slider)
“Automotive safety and fuel efficiency applications appear to be aimed at values important to consumers (personal safety and fuel economy at a time of rising fuel prices), yet these applications appear to be slow in rolling out and many are targeted only at high-end vehicles.

Will the success of this industry depend on government regulations rather than consumer demand?”

Christian (Ford) and Wiegar (Continental) described this as “A difficult question.” It was clear to me that the automotive industry is keeping a wary eye on CO2 standards (Europe) and Café standards (US) and that it would prefer to be less regulated than more regulated. It seems likely that safety applications will attract more government interest in the future. Although both Intermap and TeleAtlas were non-committal on the question, I suspect that government standards would benefit Intermap and TeleAtlas, as well as NAVTEQ, especially if safety applications requiring 3D map databases were to be mandated at the federal level in the EU and US.

Question 5 (A Spitball)
“TeleAtlas has announced that 1.25 million corrections to their most recent MultiNet (2009.09) database were based on crowd-sourced information derived from TomTom’s MapShare or from the use of the company’s PNDs functioning as probes.

In addition, OpenStreetMap is attempting to build navigation quality databases for many countries in the world based on crowd-sourced information

Google’s new navigable map data for the United Statses are based, in part, on crowd sourced data.

Do you have an opinion on the use of crowd-sourced map data and its potential contribution to automotive safety and fuel efficiency applications?”

I started the responses with Chris Wilson of TeleAtlas, who is a huge proponent of the use of crowd sourced data (especially probe data) in creating not only navigation databases but for adding ADAS-quality attribute data to navigation databases. Chris did not disappoint and described some of the areas where he thought User Generated Content could be of great value. As I expected, Eric DesRoche of Intermap challenged Chris by indicating that he could not see how crowd sourced data could be used to meet the accuracy standards required for automotive Safety or the slope data required for energy management applications.

I was surprised that Christian Ress, Ford, was not supportive of the concept of crowd sourced data. He, and other members of the panel (with the exception of Chris Wilson), raised issues concerning the accuracy, reliability, coverage and ownership of crowd sourced data and how these were problems for use in advanced applications focused on automotive safety and energy management.

I, too, have my doubts about using crowd sourced data for safety applications requiring high accuracy geometry and spoke to Chris after the session about this concern. Chris was very helpful and indicated that the probe data being used by TeleAtlas avoids many of the questions raised by the panel. In addition, he felt that some (but not all safety applications) could be assisted through the use of databases built on top of probe data. We agreed to disagree about whether UGC could provide the coverage and accuracy required for the advanced automotive safety and energy management applications. I think we will need to visit this topic in the near future.

One of the important issues that we did not discuss during the session was the strategy of map vendors partnering to meet the stated needs of the automotive industry. It is clear to me, based on what I heard the vendors say in our session at TeleMatics Update that they want the data now! In addition, the representatives from the automotive world were saying “Stop making this so hard. Find a way to cooperate, conflate your data, and let us get on with creating advanced safety applications.”

It appears that neither NAVTEQ nor TeleAtlas currently have the technology to create the comprehensive 3D Geometry databases needed for automotive safety and energy management applications, but they do have highly attributed navigation data. Conversely, Intermap seems to be creating the required high accuracy 3D Geometry data at a much faster pace than either NAVTEQ or TeleAtlas. On the other hand, Intermap, appears not to have all of the road attribute data that are needed by the automotive industry to create comprehensive solutions for all automotive safety and energy management applications.

It seems to me that there should be some partnering discussions going on here and it is clear that this is exactly what the automotive vendors were telling the map vendors at the conference. Obviously, TeleAtlas and NAVTEQ are not going to partner. Presuming that Intermap’s data passes the accuracy test, the real question seems to be “Will TeleAtlas or NAVTEQ choose to partner with Intermap?”

Of course, since this is a table-turning type of year, perhaps Google might partner with Intermap to further its interests in navigation?

Google, by the way, was at the TeleMatics Update Conference (as was Yahoo, but not Microsoft) and I now know that Google has an Automotive Division responsible for delivering cloud-based services to the automobile. Google’s goal is to get their apps in the car, but not to be involved in their design (describing that as the purview of the manufacturers). I am sure this came as a disappointment to the major automobile manufacturers, many of whom seem interested in creating a walled-garden within their cars, by controlling connectivity. Some people never learn! Hope the map vendors do!

Click for contact Information for TeleMapics

Bookmark and Share

Posted in ADAS, crowdsourced map data, Google, Intermap, Mapping, Mike Dobson, Navteq, Nokia, routing and navigation, TeleAtlas, User Generated Content

(comments are closed).