« Volunteer Register | Main | Consolidated Hardware Clearing House »

UK Open Mapping

FROM: Ben Daglish

What NEED does this meet?
A big bugbear in the UK GIS community is the fact that the Ordnance Survey (the main source of accurate UK map data) holds and enforces copyright to all its products (although it's paid for by our taxes...). Even if you just want a simple "our location" map on your website (or business card, or party invitation), you have to either pay somebody for the privilege, or go out and survey it yourself. I'm proposing an attempt to put together a high-resolution alternative source of UK data, open to all. As a side effect, other data (postcodes, road names, routes etc.) could also be collated and made available.

What is the APPROACH?
The basic plan is a collaborative site. As GPS systems grow more common (in-car, mobile phones etc.), it is hoped that most streets in the UK could be mapped by their residents - stand at one end, note the co-ordinates, walk to the other end, do the same, then come home and type them in. Schools etc.could become involved, surveying and mapping their own immediate areas. Going out for a drive? Keep the kids from getting bored by having them note down road names and co-ordinates.
A good starting point (to 'seed' the DB) would be the Digital Chart of the World data. This, although low-res, contains much of the basic information (major roads / towns etc) needed to 'marshal' the collected data.
Pre-copyright (> 50 years) OS maps, if hunted down, could also be used for starting-off purposes - although most major cities mapping will obviously have changed greatly, many more remote places may still be exactly the same, and terrain details should, in many cases, be exactly the same.

What are the BENEFITS to people?
Free (today the UK - tomorrow the World!) maps for all - created by the people, for the people :)

What is the COMPETITION?
Nothing I know of. There are a few online pre-copyright OS maps (often for model railway enthusiasts!), but nothing with this scale, and certainly not collaborative.

What BUDGET & LOGISTICS are required?
Not much :) As a Perl/DB/Web guy myself, I reckon that it'd take 3-4 months to build - certainly by that time it should be ready to receive data, cross reference for accuracy and produce basic maps. As to how long it'd take to 'finish'...well, we all know how intangible *that* is :) This obviously would be an ongoing project, but once the basic mechanism is there, it should be 'self-running' - further work would simply be in enhancement of presentation,efficiency etc.

October 31, 2003 in Geographic, User Created Content | Permalink


This is much the same as 'collaborative cartography system', 4 entries earlier in the list. Perhaps the fact that it has been suggested twice means there is a need.

I'm certainly all in favour of this idea, and would be happy to do some of my streets with my GPS. I also have a lot of existing mapping data, but it's of caves which is somewhat tangential to the primary surface-mapping ideas here.

Are there open standards for GIS information that can be used for all this stuff? The resulting database could be quite large which has resource implications.

I do think this could work, especially if it starts off with a farily specific target of mapping all the roads, footpaths, cycle tracks, houses etc. There do need to be sufficiently easy ways to get accurate info in - I imagine some tricky interface design (how do I import all the info on a pre-copyright map sheet for example?)

Posted by: wookey at Oct 31, 2003 1:25:44 PM

This is a seriously hard thing to do purely because of the numbers involved.

There are something like 250000 miles of roads in the uk. If this is done on a voluntary basis (say 3 hours, 2 days/week and you want a complete map within 2 years then you're looking at a well organised and motivated team of nearly 150 surveyors and someone to manage their work. In addition you'll need to check the data consistency, manage a _huge_ database...in short duplicating a lot of the Ordnance Survey's work.

I would argue that it may be more worthwhile to get the OS to provide some sort of subset of their (our!) data on a non-commercial license which could be hosted and accessed via a web service with an open API for non-commercial projects.

Posted by: andrew at Oct 31, 2003 1:40:43 PM

Yes there are plenty of open standards for this sort of stuff.
GML is an xml based format for describing geographical data and can be imported/exported to/from most free and commercial GIS.
PostGIS is an add-on to PostgreSQL for spatial data, although I'm not sure how it'll scale to nationwide street level map data.
There's also UMN map server for seving map data up via http.
Google and ye shall find them.

One other thing worth considering are OD heights. I can think of a few instances where a third dimension would be useful. Some thought would need to be put into how to calibrate different surveyors altitudinal measurements.

Posted by: andrew at Oct 31, 2003 1:55:31 PM

I think the important point with open mapping is that there is a license that ensures freedom on the maps' usage, basically copyleft. Why not allow commerical usage, as long as any updates the company makes have to be released publically? 150 people doesn't sound that much to me, I thought it would be more.

Have you seen the getmapping.co.uk site? It sells aerial photos of areas and I think they iron out the non-linear effects pretty well. Basically, find a few known points on one of their 2km x 2km maps and you just need an application to click and describe the landmarks on it, with the application working out the lat/long for you from the known points. Just an idea.

I've tried it with a GPS and an aerial map for my local area. I guess the big question is what language to use to describe the landmarks? How do you say a road is one way? Do you want to store a shop's opening hours?

I think the key things are
1) The right licensing for the maps and
2) The right open language for describing them.

And, oh yes, 150 people. But good idea, anyway.

Posted by: D Benson at Oct 31, 2003 1:56:28 PM

I'm not a techie, but I think this a great idea!

Posted by: Phil Green at Oct 31, 2003 2:57:07 PM

Have the same idea - registered http://www.freeworldgazetteer.org.uk/ a while ago...

Cheers Daniel

Posted by: Daniel Harris at Nov 1, 2003 9:27:24 AM

"andrew" wrote:
> ...manage a _huge_ database...

I wrote the data preparation program for the
Ordnance Survey Interactive Atlas of Great
Britain CD, did the preparation for the first
edition and helped with the second and third

The input and output data filled a significant
proportion of a CD and the preparation process
took a large part of a day on a 133MHz Pentium so
it's not a trivial amount of data but hardly
"huge" - particularly as Moore's Law has been
chomping on the problems for a couple of years

Posted by: Ed Davies at Nov 2, 2003 3:42:24 PM

"Edd" wrote:
133MHz Pentium so
it's not a trivial amount of data but hardly
"huge" - particularly as Moore's Law has been
chomping on the problems for a couple of years

Raster data such as the Interactive Atlas is trivial to handle, but we're talking vector data here with multiple attributes, topologies etc. It _is_ a huge task, a raw processing power is only a small part of that. The ability to manage 500 or so people to produce consistent, high quality data will be hard.

Posted by: andrew at Nov 4, 2003 12:08:40 PM

There is also a patent issue vis-a-vis solutions which will use the Open Mapping/Cartography system.


The European patent EP0845124B, entitled "Computer System for Identifying Local Resources and Method Therefor," relates to Multimap.com's technology for displaying both an image of a map and information data relating to at least one place of interest on the map to an end user

Posted by: andrew at Nov 4, 2003 12:16:14 PM

andrew wrote:

>> Raster data such as the Interactive Atlas is trivial to handle, but we're talking vector data here with multiple attributes, topologies etc.

Actually, the OS IAGB is/was vector data (at least the first three editions were - I haven't taken much notice since then).

andrew also wrote:

>> The ability to manage 500 or so people to produce consistent, high quality data will be hard.

Absolutely agree - the data size is not the problem as such but the consistent management of the large number of sources with different interests would be.

Posted by: Ed Davies at Nov 4, 2003 3:34:31 PM

We have a site that links locations at www.gole.co.uk. We are currently using it as a blogfinder. The simplest description of the site is - a searchable map allowing the user to find interesting urls by location. Locations are placed by postcode, utm grid, or lat and long. We also have a trial method of placing a spot on the map by "eye".

I would appreciate any feedback offered.

Posted by: Robin Cousins at Nov 14, 2003 7:34:57 PM

Great idea. If speed limits were added this could be used with a GPS device to be a speed warning device - I want one to 'nanny' me when driving as I am terrified of getting more points and do not want to exceed the speed limits, but no device currently offers full-time warnings.

Posted by: Rob at Nov 21, 2003 12:19:58 PM

>>do not want to exceed the speed limits

Don't worry, if you don't exceed the speed limits you won't get caught.

Posted by: Andrew at Nov 28, 2003 3:30:59 PM

We've been working on something similar for a while - http://www.geowiki.com/ . Major map update currently in progress.

Posted by: Richard Fairhurst at Sep 12, 2004 4:17:25 PM