[Talk-GB] Rebooting the NAPTAN import?

[SK53 on OSM]

On 02/04/2011 15:14, Stuart Grimshaw wrote:
> This weekend sees Sheffield's first Transport Hackday, hosted at
> theGistLab[3], we've got a bunch of guys sitting here with various
> apps we've written or would like to write and one of the projects we
> identified as a potential target for today would be refreshing the
> NAPTAN data in OSM.
>
> Myself&  one of the other hackers have been reading the info on the
> existing import[1] and we've come across a few speed bumps in our
> efforts.
>
> Firstly, the info on the page is about 2 years out of date, so my
> first question is what is the state of NAPTAN data in OSM? Did the
> import finish?
>
> Secondly, the license mentioned seems out of date, since 2009 the
> NAPTAN data has been released under the Open Government License which
> makes using it with OSM much more palatable[2]
>
> I've seen the tools available to import NAPTAN data, and now that the
> license is relaxed, I'd like to propose we kick off the data import
> again?
>
>
> [1] http://wiki.openstreetmap.org/wiki/NaPTAN/Import
> [2] http://www.dft.gov.uk/naptan/termsOfUse.htm
> [3] http://opendata.thegisthub.net/2011/03/transport-hack-day-datasets/
>
> -S
>
> Follow me on Twitter: http://twitter.com/stubbs
> Blog: http://stubblog.wordpress.com
> My art: http://stuartgrimshaw.imagekind.com
> Stock Images: http://en.fotolia.com/partner/16775
>
> _______________________________________________
> Talk-GB mailing list
> Talk-GB at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk-gb

I would suggest the following:

a) Do not import any NaPTAN data in areas where imports have already 
taken place. Experience shows that detailed survey & correction of 
NaPTAN data is not to be undertaken lightly. IIRC about 10% is wrong. 
The best data are for Hullwhere Chris Hill surveyed the lot. I have done 
only about 20% of Nottingham's NaPTAN stops and have a similar error 
rate. Unfortunately processing NaPTAN alongside primary surveying just 
didnt prove viable, but there are plenty of stops which no longer exist, 
have moved or dont exist on the ground.
b) Check with any mappers in the area before performing an import. There 
may be good reasons why they have not requested one in the past.
c) The best approach would be to host current NaPTAN data in a location 
where OSM data can be compared & then mappers could choose to import it. 
Having an application which did this would be way more useful than 
shoehorning NaPTAN data in on its own.
d) Bus stops which have disappeared are difficult beasts. Have they been 
accidentally deleted, or has a mapper actually surveyed a site & deleted 
the stop because it doesn't exist.
e) I have literally hundreds of bus stop waypoints which I have never 
got round to cross-checking against NaPTAN. Others may be in a similar 
position. You can see this where a NaPTAN stop & a previously mapped 
stop exist close together. An import will either sit in the database 
gathering dust, or it will impose a substantial workload on local 
mappers if they want to check it. We dont yet have enough mappers to 
cross check this kind of imported data in a jiffy.

For me the NaPTAN data was most useful for naming roads & spotting 
places which needed a bit of TLC. I use bus stops on the Garmin so I 
find the data useful, but I would urge extreme caution about importing.

Cheers,

Jerry Clough