[Talk-transit] NaPTAN case study
Jerry Clough - OSM
sk53_osm at yahoo.co.uk
Thu Oct 22 14:45:38 BST 2009
Unlike Chris I have not been actively surveying lots of bus stops: there is too much other work in Nottingham for this to be a worthwhile activity in the short term. Also the NAPTAN import has exposed areas which were traced from Yahoo: in verifying one stop I had to move a considerable number of ways (something I knew needed doing, but had been putting off). My general impression (probably based on around a 10% sample) is that most Nottingham data is well located: I'd be surprised if real changes is needed for more than 5% of stops. The major error detected so far is on the site of City Hospital where a bus stop is shown in the middle of the Urology building. I suspect that there are a number of other stops which have been relocated in this area as well, but it requires another on-the-ground survey.
However, I have been using the data (as I think I've already posted) for naming and adding roads. Of course some of this may generate some errors, but probably no more than those from transcribing dictaphone surveys (names one road out, or on opposite side of street).
I don't have the precise figures, but I think I've added about 50 roads purely based on lines of bus stops in poorly mapped towns (Bolton, Oldham, Mansfield, Grimsby). Others have used the presence of NaPTAN data to assure continued existence of a road since NPE.
I have used NaPTAN data to name substantial numbers of roads: particularly in Nottingham, Wirral, Oldham and Grimsby areas, but also in smaller towns and villages.
NaPTAN data is also useful for checking boundaries of authorities. For instance the OSM boundary between Notts & Derbys is shown to be inaccurate by NaPTAN in the area of Pleasley : http://osm.org/go/eu9D9Pfu (easy to see because no import for Derbyshire)
Using NaPTAN data in this way gives one quite an insight as to the overall quality , adherence to guidelines etc. My general impression is that the Merseyside & Greater Manchester data is less useful than for other areas (Nottingham, Notts, Lincs, NE Lincs etc): most stops have names in the form of Major Road/Minor Road, the indicator field is often inaccurate or uninformative (patch of waste land in Croxteth for instance) and elsewhere on Merseyside some data appears to have been copied from stop to stop. To a certain extent the overall usefulness of the data for other purposes in OSM depends not just on the quality of the data, but also the degree to which the route network penetrates into estates. Nottingham has quite a number of routes which do this using smaller buses. It is really noticeable that quite large areas of Merseyside are only served by buses running along main roads.
Incidentally the data for the Wirral also includes some taxi ranks.
SK53 on OSM
From: Peter Miller <peter.miller at itoworld.com>
To: Public transport/transit/shared taxi related topics <talk-transit at openstreetmap.org>
Sent: Thursday, 22 October, 2009 13:54:13
Subject: Re: [Talk-transit] NaPTAN case study
On 22 Oct 2009, at 12:55, Chris Hill wrote:
> Christopher Osborne wrote:
>> Hello all
>> As part of the data.gov.uk <http://data.gov.uk> experiments, I had
>> several encounters with top brass of various gov departments. They
>> were very excited in the NaPTAN import, none of them had heard about
>> it and as far as I know it is the first example of crowdsourced
>> improvements to a UK gov dataset.
>> Very glad to see http://mappa-mercia.org/novam/ is back online today.
>> I need to prepare some briefing information to send back to top
>> and was wondering where are some good locations to show off the great
>> work OSMappers have been doing with the NaPTAN data?
>> Christopher Osborne
>> www.itoworld.com <http://www.itoworld.com>
> We have checked about 75% of the stops in Hull. I have established a
> contact in the city council's transport team who has received the data
> we have checked with all of our comments about what we found so
> far. I
> hope he will use the data to feed back to NaPTAN so eventually the
> quality will improve all round.
> I created a simple overlay for the city http://bus.raggedred.net to
> to show the progress but also to highlight stops that we have found
> need correction to the NaPTAN data. I like the idea and the look of
> NOVAM, esp. because it is national and updates regularly, but it
> yet show quite what I need to help the council team.
In Ipswich I have re-surveyed the NaPTAN data for about 1/3rd of the
town so far and found errors in about 5% to 10% of the stops
(including major transport interchanges such as the bus station,
railway station and hospital) which have all been reported and most
have now been fixed in the source data. I will complete the survey
when NOVAM is up-and-running.
Just to note that Chris and I had a meeting with Transport Direct
earlier in the week and we were discussing ways to open up a direct
reporting system for faults back to the relevant data owners with some
layers of management reporting back to the DfT to ensure that reports
are dealt with. This should help with feedback in time, more on that
in due course.
I also had a review meeting of the Ideas in Transit research project
yesterday with the Technology Strategy Board and they were keen to
know how the NaPTAN data was being used within OSM to do interesting
things and wanted me to put a report together about what was
happening. I mentioned that the data now appeared on the main OSM map
renderings. Anything else I should mention?
In summary, I think that this import is now being seen by the UK
government as possibly the first major example of crowd-sourcing for
the improvement of official data, and that if they can show that it
has been successful then more data may follow. :)
> Cheers, Chris
> Talk-transit mailing list
> Talk-transit at openstreetmap.org
Talk-transit mailing list
Talk-transit at openstreetmap.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Talk-transit