[OSM-talk] Fwd: Nav4All navigation shut down by Navteq

Kai Krueger kakrueger at gmail.com
Mon Feb 1 17:58:09 GMT 2010


On 01/-10/-28163 08:59 PM, NopMap wrote:
>
>
> Hi!
>
>
> Kai Krueger wrote:
>>
>> Even if promoting to the company did not work, it might still be worth
>> promoting OSM in the media coverage / blogs of the Nav4All shutdown as a
>> reason why it is important to have free and open map data to prevent
>> anti-competitive activities by a few large companies.
>>
>
> I am fully aware that many people will not want to hear that - but why
> promote something that OSM cannot deliver? I can understand why Nav4All will
> rather shutdown than attempt to switch to OSM.

I'd agree with you if you'd insert a "OSM cannot _yet_ deliver". No, I 
don't think OSM can currently quite compare with the full Navteq on a 
global scale in those countries that Navteq has good coverage for 
turn-by-turn applications. And I would probably even go as far as say I 
can also understand why they would rather shutdown then currently use 
OSM depending on where their main user base is. But that doesn't mean we 
can't use the fact that a company (potentially) has to shutdown due to 
licensing dispute between competitors, to make the point why it is 
important to have open data and try and convince people that is worth 
their while to contribute to such an effort to ensure that it does 
become a viable alternative. And I am convinced that OSM will continue 
to become viable in more and more applications including eventually 
turn-by-turn navigation.

>
> A while ago, I had approached another company which produces mobile
> navigation software, where I know some people. I tried to advertise OSM data
> and maybe get some support for their software. To my great surprise, they
> had spent a considerable sum of money on converting OSM data into their
> format and had already evaluated it. The result of the evaluation was
> disillusioning: The geometry is pretty good, but the attribution is way
> below what would be required to substitute the commercial data. They decided
> to not use it, in spite of the work already invested.

Interesting. I think that could also be "spun" positively ;-) It means 
people in the industry are starting to take OSM seriously and actually 
invest money to evaluate how far it has come and be prepared for when it 
does reach a sufficient quality or need to quickly switch. It also means 
they must have had some confidence in that the process of crowd sourcing 
map data can work. Again I would agree with you that geometry is good 
and attribution still somewhat lacking. Osm is missing loads of turn 
restriction, height or weight restrictions, speed restrictions and 
housenumbers to name a few, even in areas with very good geometry 
coverage. But from a point of view of being disillusioned, I think in 
the majority of cases they are missing and seldomly wrong. So it just 
needs a lot more mappers and some time and that should be achievable too.
Without knowing the company and any more of what they concluded I 
obviously can't say if the above statement is true for your example. But 
I have at least been peripherally involved with writing the turn-by-turn 
  routing support of GpsMid that is based on OSM data and in my limited 
testing, the routes it found in high coverage areas, were not really 
worse than those found by a TomTom or Navigon that I had for comparison. 
Each had parts where it was better and worse than the others. So I do 
think it would be possible to make good routing from OSM, given good 
(commercial?) software.




>
> Which is rather close to the statement from Nav4All.
>
> OSM is quite suitable for any hobby project, but I believe that the
> anarchistic nature and the often controversial and sometimes disputed and
> chaotic tagging are reason enough to deter the use of OSM in any
> professional area where you are talking about warranties. The commercial map
> data has fixed tagging schemes and minimum quality standards. It contains no
> nasty surprises in general and if it does fail in some places, there's a
> provider who is liable to fix this ASAP. As long as OSM has no comparable
> standards (and I don't expect it will have - I'd like to point at my
> favorite example that there's still no agreed way to tag something as simple
> as a bicycle way), it is unlikely to meet the existing standards of
> commercial providers.

So as I stated above, I don't think the _main_ problem at the moment is 
the anarchistic tagging, but still too limited coverage, especially on 
tagging relevant for routing. You just need to look at the Navteq and 
Teleatlas maps and see how many errors they contain and particularly how 
outdated in many areas, even on major roads in the middle of major 
cities they are, that you must realize that commercial companies can and 
have to be able to live with errors, inaccuracies and "nasty surprises". 
And companies aren't liable for the errors, I don't think, or has any 
company been successfully sued for drivers driving into railway tracks, 
into rivers, getting stuck on unpassable roads as one can occasionally 
read in the media, happens?

>
> I am aware that other than Nav4All and the company I talked to, Skobbler is
> trying to switch to OSM. They are probably running into all the problems
> with ambiguities and controversial tagging right now. So I am very
> interested what sort of navigation they will manage.
>
> bye
>          Nop
>

Kai





More information about the talk mailing list