Category: Thoughts

I’ve have been at the Ordnance Survey 5 years now.. and at times I would be the first to admit that life has often been very frustrating… The organisation has come on leaps and bounds in many ways over this time, but I still feel like the proverbial “captain of a oil tanker” trying to turn his ship.

So to vent my frustration to some extent, I present today the first of a number of irregular postings which describe my own personal view as to what the OS needs to do differently, if the OS does not manage to get there, well at least you will understand my intentions – and hey I have got these off my chest.

Let me make it clear at the start, the activities of the OS are highly valuable, and as I have made clear on many occasions I believe the operations of the Ordnance Survey (or any mapping agency) are best funded directly by its users through licensing data rather than general taxation.

There is however, an argument, that the requirement to license data restricts it’s use, and in particular, reduces innovative uses of geospatial data. I don’t think however the problem here is actually licensing data for specific activities, I believe the issue is that the OS makes it too difficult for potential users to understand the value in its data.

Historically the OS has made it very difficult for anybody to gain access to its data, in many markets there was no alternative to using Ordnance Survey mapping and very restrictive controls of it’s use were enforced (When I joined the OS it was not possible to license OS data for display on the web in any from whatsoever !!).

Well clearly the Landscape has changed..

There is now real choice for many users of OS geospatial products, both in terms of other commercial providers and through the developing open source geodata movement. Also, as with the general software industry, the customer is taking control, demanding much more from their providers – wanting better understanding of their business, solutions rather than pre-packed solutions etc.

The software industry is responding to these demands by changing the way it operates, and the OS needs to the follow these trends.

Software publishers, like geospatial data providers have traditionally exploited IP in the form of packaged solutions, this is changing – increasingly software is becoming a service business, where in the past corporate customers might have bought a CRM system, now they rent usage at salesforce.com.

Geospatial data on its own can never deliver a total solution in the way that salesforce.com does, however the OS with it partners can provide geospatial data as a managed service… do customers really want the hassle of installing GIS software, loading and converting data at regular intervals ?

Providing OS data as a service to an increasingly networked group of potential customers massively reduces the barrier of entry to geospatial information.

The big difference I think this would also make, is that is would allow potential users to “Try OS data for Free” – OK pick yourself up off the floor and let me explain..

Let users discover the value in OS data by actually deploying it, and if the value is there, they will pay for it later.
This is clearly based on trust, but the reality of our industry is that for too long we have focused on the buyers of geospatial information and not its users, who are often the innovators but who need access to OS data for example, to figure out its value.

Most enterprises are honest and when a real need is identified and a solution found they will pay!! This is how salesforce.com has become successful selling to sales guys who were frustrated by the big iron ERP systems, it is also behind the success of the Blackberry which was a personal buy before it became corporate.

So Ordnance Survey 2.0 must work at reducing the resistance to identifying the value in geospatial data by concentrating on the delivery of information within a service and by making the service initially free – if the user sees no value then simply switch the service off.

Written and submitted from the Holiday Inn Express Southampton, using my Vodafone 3G network card.

I blogged a few weeks ago about the generally unrecognised potential value of geospatial “attention data”, the information which describes what we are interested in, as manifested by the information we leave behind when we use web-sites.

The BBC today reports on Googles harvesting of its attention data using its new Trends analytic tools.

It appears that Birmingham in the UK is the centre globally for searching for porn on the internet, following by Brisbane, Perth and Sydney and Melbourne – does this say something about Australians? Very bizarrely these major cities are followed by Brentford a suburb of London just a few Km north of Teddington…

I smell something fishy 🙂 with these findings, but regardless of how this information may have been analysed, I think it still demonstrates the value of this type of data.

Written and submitted from home, using my home 802.11 network.

I’ve just got back from the inaugural meeting of the British Computer Society’s Geospatial Interest Group, another example of GI hitting the mainstream. I was asked a number of questions on positioning technology and the accuracy of data, which reminded me that the OS has just completed it’s Positional Accuracy Improvement (PAI) programme.

Like most other national mapping agencies and other creators of national geospatial data products, the data contained in the OS National Database dates from a time before absolute positioning technology, where the location of a feature was ‘measured’ relative to a fixed control network and other “stable” geographical features.

For most purposes this has not been a problem, however with the introduction of GPS and the fact that the data was originally collected using local country projection systems, random location errors of up to 10 metres became obvious in rural areas.

The OS was left with the dilemma, ignore the issue – or correct it. This is not as obvious a decision as you might expect, in the UK most organisation who had captured they own information had done so in relation to features as positioned on OS maps, and these features might now be moving.

There really was no choice for the OS, as with many other organisations the widespread use of GPS meant the data has to be corrected, so five years ago the project was started and it was completed last month !

Out of the 244,000 sq Km of Great Britain, over 160,000 sq km had to be corrected using photogrammetric survey to increase positional accuracy to around a metre in most cases. To allow user defined data to be adjusted geometrically, no less than 450 million change vectors have been calculated for use in transformation software.

The OS may be one of the first organisation to address this issue, but I’m sure many other organisations will have to follow the example, the relative accuracy’s of mapping data used in the early days of digital cartography are no longer ‘fit for purpose” in the era of widespread GPS use, so for many people the earth may well be moving!

Written and submitted from home, using my home 802.11 network.

Following a comment by Charles Arthur (yes he of the Free Our Data Campaign) that there were no UK based mapping mashups, I did a bit of a search and came up with this page which lists at least 22 mashups in the UK, and yes they all contain Ordnance Survey data licensed through partners.

If you know of any more I’d love to know..

Written and submitted from the Holiday Inn Express Southampton, using my Vodafone 3G network card.