Thursday, 22 October 2015

A visit to the Ordnance Survey HQ in Southampton

For a while in the early 2000s I was a member of the Charles Close Society, a learned society for anyone keen on the Ordnance Survey and its maps, whether as an amateur or as a professional. I wrote two articles for their journal, Sheetlines. This is the first. It describes a visit I long wanted to make, to the OS's headquarters at Maybush in Southampton, where they had been based for many years since moving out of their former city-centre buildings. I saw the Maybush site in November 2001. More recently, the OS has moved to brand new premises (still in north-west Southampton) much more suited to producing mapping in the Digital Age.

I didn't stay with the Charles Close Society, finding it just a little too dry and academic at the time. But it's a most worthy organisation. I may rejoin one day. Their website is here:

Here's the 3,714-word article as submitted. The published article in Sheetlines (April 2002) was only slightly shorter (the lunchtime observations were shorn). Writing it was not originally in my mind. The organiser of the visit, Gerry Jarvis, asked me to produce a typed record of the visit, as one of the things new Charles Close Society members took their turn at doing. I was determined to make a very good job of it.

I've added some photos I took at the time, even though they did not form part of the published article.

Vanessa Lawrence, by the way, was the first woman to hold a prestigious ultra-high-paid appointment as head of a modern UK government agency. She's in Wikipedia at


The Ordnance Survey in Transition

This visit revealed something of the modern Ordnance Survey, steered since 2000 by its new Director General and Chief Executive Vanessa Lawrence.  We were shown the first fruit of her leadership.  It seems we may expect some pretty radical changes. The CCS visit organiser Gerry Jarvis deserves a great deal of thanks for arranging such an eye-opening day. In the course of five hours, we had a glimpse of the OS's latest business initiatives, survey methods and mapping. The focus is now even more on the business customer, but the general public are not forgotten: indeed, the OS wants to draw in a wider range of users. We were the particular target for the day, but there was no hard sell. It was all very enjoyable, a good mix of experiences; and we were welcomed, ushered, informed, entertained, and given refreshments by a very friendly and professional team of OS staff. We were in fact treated like business guests who might become important customers. And many of us partly justified these hopes by buying a large number of maps at the Map Shop at lunchtime!

The Ordnance Survey has clearly entered an important new era, committing itself to becoming the dominant provider of computer-based maps, downloadable on demand by users who can select the data they want to see.  This means assertive - perhaps aggressive - development and marketing, with as much as possible done to promote the database and win new customers. The OS wants to make many of the map-illiterate familiar with maps and their uses, children not excepted. So there is an educational thrust here as well. The OS is out to sell and popularise its products to everyone, and in doing so corner the market. Money is a huge issue. Defining the product; targeting the customers; decisions on appropriate quality; cost control; return on capital; these business considerations will probably now drive its future course ruthlessly. One might ask, what is the economic working life of its printing presses? Will the OS still be making and selling traditional paper maps in ten years time, or will it be relying wholly on across-the-board electronic delivery? The Rambler for Winter 2001 does in fact mention 'a trial OS project at the New Forest Visitor Centre in Lyndhurst [that] allows visitors to get instant print-outs of maps from a coin-operated touch screen terminal. Users can search by postcode or place name before printing out colour paper copies for their chosen walk’. This is only one possible outcome of current thinking. It seems that the days of the large, folded paper map with a cover may be numbered. And also the idea of fixed sheetlines. The passing of maps as documents may not please everyone, and in the short term the OS is not going to risk a fall in revenue by jeopardising its strong paper map sales to the general public; but if electronic maps catch on, there is bound to be more of the same.

Arrival, and the events of the morning
The security guard knew all about the CCS visit, gave clear directions, and added, ‘Nice to meet you’.  It was an efficient, welcoming start: the OS's charm offensive was beginning at the gate! Money had been spent on making the public entry point look modern and inviting. Inside, there was a completely refurbished and attractive interior, with much light-coloured wood, and a sea of blue carpet, lit up by spotlights everywhere. Facing you was a circular reception enclosure, rather like an atoll in a tropical sea. It was staffed by a series of pretty, smiling girls. It could have been the entrance to a new hotel. Whatever drab corridors lay out of sight on upstairs floors, here at least the OS was presenting an image designed to impress the visitor. Not far off was an illuminated notice, which said, 'ORDNANCE SURVEY WELCOMES MEMBERS OF THE CHARLES CLOSE SOCIETY - WE HOPE YOU WILL ENJOY YOUR VISIT'.

By the entrance was a discussion room with walls of glass, and within, as if intended to be seen by visitors, were several people holding a conference. Perhaps it was just the Canteen Committee; but they had the dress and demeanour of accountants, and perhaps the fate of the Landranger series was at stake. But off to the left was the Map Shop. Rank upon rank of orange, yellow and magenta covers on spotlit pine shelves, apparently the entire published output, pristine, waiting, and the shop not open until lunchtime! One fluttered like a moth outside that locked glass door.

Gerry made himself known, warmly welcoming a party of eighteen, ten of whom had never visited the Southampton site before. Our Ordnance Survey hosts Alan Sivier and Angus Tweedale had been waiting in the wings, and promptly at 10.30am led us into the new Ordnance Survey Business Centre, opened only five months previously.

This had an impressive entrance of its own in stainless steel, also spotlit, and was a large, self-contained inner building. All around were areas where business conferences and exhibitions could take place. The lighting was theatrical, spotlights aimed at ceiling panels which bounced the light back onto tables, leaving wall-mounted flat TV screens shaded for presentations. Several meetings could have been held at once in this open-plan space without interfering with each other. We proceeded to the Garden Room.

This was a large, sunny refreshment room, where coffee and biscuits were laid on. It was an opportunity for members to introduce themselves properly, and describe their journeys - for many had travelled a long way to attend. Then at 10.50am we adjourned to one of the Conference Rooms. This was just a room with a projection screen on one wall and plenty of seats: it was going to be our base for the day.

At 10.55am Ian Evans, the OS's e-strategy Manager, began a presentation using the computerised projector about how the OS was transforming itself into an e-business. He explained how the The Vision to 2005 had been introduced by Vanessa Lawrence. The OS was to be 'the content  provider of choice for location-based information' in the new information economy. There were to be five distinct strategic initiatives, and these represented the biggest investment ever made by the OS. The intention was to grow and hold off the competition, and there was no secret about it - the e-business strategy was explained in full on Ian went through some of the projects, such as:

• OS Websites individually tailored for each major customer.  These would show only the topics that the customer was especially interested in. Fifteen of these had been set up since commencement in June 2001.

• A redesign of the main OS website, with simplicity as the goal. There had been two redesigns in the last year, and another was coming that very month.

OS Direct - supplying map data online to users of the new MasterMap. Sending information by post (a CD in a packet) was prone to trouble and some delay; now major users of MasterMap data will be able to define what they want over the internet, and get it instantly online. They will even be able to specify fresh data only a day old. Early users will testbed a sample map, which will help the OS determine future customer requirements.

OSmaps24 - maps online for the general public, to be displayed on a variety of devices. This will mean instant downloads on the day, at 1:50,000 and 1:25,000 scales, for walkers and other general public users. The 'device' might for example be a kiosk, charging a fee. They were experimenting with the level of fee - £1.00 might be reasonable; the service might even be free; but there must be no risk to ordinary retail sales.

• The New OS: new ways of working. Basically an internal initiative. The aim here was to streamline administrative procedures, provide online forms procurement, online interruptable training, and new methods of ‘knowledge management’, all with savings and better efficiency in mind.

• Developing the market - promoting the OS brands. This was all to do with imaginative branding, quality standards, pricing policies, and licensing users.

Our attention was now drawn to a table at the back of the room, where various freebies were laid out. These included the 2001 Business Catalogue, a CD showing the UK from satellites, and vouchers entitling us to 25% off any purchases that day from the Map Shop. There were no spare vouchers after the rush.

Then at 11.30am it was the turn of Kathrina Ayling.

She had been a surveyor, but was now a full-time training officer. Again using the computerised projector, she gave us an on-screen demonstration of what the latest software could do. The program used was PRISM - standing for Portable Revision and Integrated Survey Model. Using her laptop, Kathrina showed us how to plot a building onto a blank area of plan, starting with an accurately-surveyed base line.

It was fascinating to watch points being plotted onto the plan, and then being joined up with coloured lines, representing the sides of the building or fences, all with just a mouse click. Curves required a minimum of three points, but straight lines only one if measured from an already-plotted base line. It did require care and dexterity, but the process was very rapid. Many fewer measurements were needed, and the program automatically tested whether the data plotted was within tolerance - whether it would 'fit' what was already on the underlying map. No fit, no plot, and the surveyor would have to do it again.

With her was a trained surveyor called Mark who demonstrated the hardware surveyors were now using. First, Kathrina passed around the flat, portable Fujitsu computer which the surveyor carried - basically just a tablet-shaped screen for plotting on with an electronic pen. It had a Pentium 3 processor and a 15 gigabyte memory. So it was fast, and capable of storing a great deal of information. Without a cover it weighed about a pound, and the dimensions were roughly 12'' long by 9'' wide, and 1'' thick. But it was used in a blue plastic showerproof case, and this made it heavier. That too was passed around. Then Mark put on the full gear.

On his back was a special backpack for the support electronics. The plotting tablet in its case had a tummy-pad and a harness which held the tablet at a convenient angle out from the body. Hooked to his belt was a rangefinding device for sending a narrow laser beam towards any solid surface in the line of sight, and accurately measuring the distance: although a difficulty with that was seeing where the red spot was on the building! His writing hand held the pen. In the other hand was a pole with a flat saucer-like top: this picked up signals from GPS satellites orbiting the Earth. Wires went from the tablet, rangefinder and pole to the backpack.

Mark explained how he would use GPS to locate exactly where he was. It was possible to get a starting position accurate to within a few centimetres. There were always four satellites somewhere overhead - four were needed to establish latitude, longitude, time and height. But the signals passed through different atmospheric conditions, and this had to be allowed for. The surveyor would add a correction, derived from two fixed bases in Great Britain with known coordinates, connected to him by a mobile phone link. Or the surveyor could determine the exact position of his car, there on site, which then became the local fixed base. Given a completely accurate starting point, the surveying could be done very quickly. Three weeks work might be done in a morning, allowing next-day availability for MasterMap users. Features plotted on the map were time-stamped, and it would be possible for users to ask for a map showing features present at any particular date. The entire kit cost about £2,500, which was not cheap, but the drastically increased productivity in trained hands justified the expense. Even so, the days of traditional surveying with yellow tripods and theodolites were not over. The old methods were still necessary if the satellite signals were poor. This was a frequent problem in cities, particularly parts of London where tall buildings hid the horizon.

At this point we adjourned for lunch.

Lunch was available at the staff canteen. This was in a separate building, and required a short stroll in the cold November air. It was a large, smart, high-class place, full of OS staff, and those CCS members who ventured inside were rather lost among the sea of faces.

Lunch over, we headed for the Map Shop, armed with our discount vouchers. It was open from 12.30pm. This shop was something of a showcase for the OS's latest publications. Besides having the full range of the most popular current series, there was other merchandise such as clothing, all bearing the OS logo, and much of it in sizes most members would have found too small! The OS was clearly looking at the under-sixteen market. Perhaps it entertained school parties, as well as members of learned societies?

The maps on display were really too much of a temptation: despite the very recent price increase, several of us found ourselves at the till with £30 worth of maps. But it was a consolation that we got that 25% discount. And there were also free OS pens and pencils with each purchase!

The events of the afternoon, including the Print Room, and then departure
At 1.25pm we were back in the Conference Room. John Walk now gave us an introduction to digital mapping, using the computerised projector.

All small scale maps were now produced digitally, and therefore one could produce a map for any area on demand, with any data desired. Small scale maps needed only raster data: that is, tiny squares, each of which had size, colour and a unique National Grid reference. Without losing position, raster data could be manipulated to show the same information at a variety of scales. Altering the colour made it possible to highlight different features of the map, or omit them. John showed us examples of the new Landplan - the replacement for the old 6'' map, which was compiled from 6'' and 1:10,000 sheets, and other sources, simplifying the detail with a computer. This had proved a very good shortcut technique, although some hand-finishing had been needed. The result was an accurate standard base map, the buildings filled with pale yellow. It was not perfect - diagonal lines had a ‘stepped’ appearance, for example - and some might deplore such departures from the traditional standards of cartography. But it was not intended for the ordinary public. Its typical use would be by local authorities and others as a robust ‘working map’ - a map you could put your coffee cup down on. It could be produced from scratch in just a few days, and was up to date - not stuck in a slow, 20 year revision cycle. Raster data also allowed 3D imaging based on contours. We were treated to an animated computer ‘flight’ over Ben Nevis, the ‘land’ below being a 1:50,000 map, enlarged and stretched into a 3D shape which mimicked the actual lie of the mountain and surrounding glens. One application of 3D imaging was to see whether some new feature could be seen from somewhere else - which might settle a planning dispute.

For large-scale plans, vector data was used. This went beyond simple coloured squares. It meant adding information of various kinds to points on the map. This allowed much more flexibilty - one could change (or omit) colours, line thicknesses and symbols, and superimpose information in as many layers as one wished. Vector data was much better for analysis. We were shown some insurance company examples, based on a Landplan of part of Southampton. Having set up all the addresses of its policyholders in the area as map coordinates, the company gave values to each coordinate. Then this data was plotted onto the map. In the example, coloured dots on each house represented existing policyholders, clearly showing up which addresses might be a source of fresh business. In another example, a range of coloured dots showed the level of claims for addresses, road by road, and this could prompt enquiry into why the claims were higher in certain roads than elsewhere on the map, leading in turn to a better estimate of risk. It was one way of matching premiums more closely to the level of claims. In yet another example, all of the map below a certain contour was coloured, representing the flooded area if the River Itchen burst its banks. From this, a list of affected addresses was easily produced. In different examples, we saw how a cable company's grid of pipes could be accurately superimposed onto a large scale plan; or how the 'best routes' for emergency vehicles could be plotted, to take into account low bridges, traffic lights, narrow lanes, and so on. There were obviously many commercial applications.

It was now 2.15pm, and we broke to visit the Print Room. This was a very different experience from the computer-based presentations seen so far. Alan Sivier led the way. We had to keep in a group - no stragglers - and not stray from a marked walkway. It was, after all, an industrial environment, and certain safety procedures had to be observed. The printing was done in a kind of large warehouse. The first thing to be seen was a bench for handling aerial photographs, which came on long spools. From these films were created, and the images on them were rectified for plane tilt.

This area was something of an antechamber to the main room, which contained two huge presses (a Nebolo Colora 7000 and a more recent KBA Rapida 130), plus a large variety of smaller machines for specialised jobs.

The older of the two presses took three hours to set up, and was therefore suitable only for large print runs. But the other, a modern, fully-computerised German machine, took only 45 minutes to set up, and could remember the settings for any particular map, enabling a short print run at any later time. It used four colours, plus an overprint colour. These leviathans seemed to be ready for a run as we arrived, but were not obviously in motion, and there was little noise. The Print Room seemed to be staffed by craftsmen going about their business quietly. Once printed, maps had to be checked and perhaps guillotined in a dedicated machine which gave new meaning to the phrase ‘cutting-edge technology’!

There were also folding machines, which could deal with printed maps at the rate of 5000 copies per hour. It was awesome to walk past waist-high stacks of unfolded maps, and contemplate the purposes of sundry pieces of apparatus that for now were not being used. Incidentally, 4,000 Landranger sheets stacked upon a pallet came to just above the knee, in case you wanted to know!

At 2.45pm we went back our Conference Room, rejoining John Walk for a final projected presentation on the MasterMap. This brand-new product was compiled using ‘area features’, meaning anything which had a shape or defining line. This included buildings, structures, parcels of land, and the entire range of land and water features both natural and man-made. All of them could be picked out or filled with colour - and also information, accessed on the computer screen with just a mouse click. The MasterMap had discarded some traditional detail, but allowed new things such as pavements and verges to be shown.

Each feature had a unique TOID (a Topographic Identifier) for unambiguous referencing and easier data exchange between users. So the Land Registry could use it in discussing a property with English Heritage. The TOID was a 16 digit number - 16 digits should be enough to accommodate every mappable feature for all time. The TOID defines not a point but a real object or area on the ground, and clicking on that feature will open a window in which will appear data contributed by users. At the start perhaps only the OS, local authorities and the Land Registry might supply information. Even so, that would include the TOID number, the grid reference, if a house the street number and postcode, details of ownership and rateable value - and so forth. The concept would take some time to bed down. As users themselves could enter data, there was theoretically scope for error, but in practice there should be little problem, as TOIDs will not be used casually, only in a computer context where a high degree of care might be expected. MasterMap was not tile-based, that is, not published only in 500m squares: the user will be able to specify any area to be shown; and ask to omit unwanted features. The OS was experimenting with new ideas for this map, such as techniques for telling the difference between various kinds of open space, or types of building.  One proposal for new housing developments, to show houses being built, was to plot information from the builder's own surveyor, tagging the site to say it had not yet been surveyed by the OS, and then going in later to carry out an accurate survey. There was more on MasterMap on the Ordnance Survey website (

By now it was 3.25pm, and we came to a finish. Gerry expressed our appreciation for a most interesting visit, and then we went again to the Garden Room for tea, coffee and an abundance of cakes before we went our various ways. It was a pleasant ending. Many thanks to the OS, and to you, Gerry.

Afterwards, the Society big cheeses were astonished at how much detail I'd noted down or remembered. But of course it was similar to the detail I'd extract during a two hour meeting with a taxpayer, in the normal course of my Inland Revenue duties at the time. After years and years of practice, I was used to memorising a huge amount of information, taking only brief handwritten notes. Every decent investigator could do this. Reading the article now, however, I am very surprised how computer-literate I was way back in 2001. Some of the other Society members there (they were mostly older men of course) had a vaguely vexed look on their faces, as if still stuck in a pre-computer world. The one or two women present seemed much more on the ball.

Anyway, I was happy with the article, and pleased that they liked it enough to publish it with hardly any changes.

1 comment:

  1. So glad to hear they wheeled out some 'pretty, smiling girls' to welcome you....


This blog is public, and I expect comments from many sources and points of view. They will be welcome if sincere, well-expressed and add something worthwhile to the post. If not, they face removal.

Ideally I want to hear from bloggers, who, like myself, are knowable as real people and can be contacted. Anyone whose identity is questionable or impossible to verify may have their comments removed. Commercially-inspired comments will certainly be deleted - I do not allow free advertising.

Whoever you are, if you wish to make a private comment, rather than a public one, then do consider emailing me - see my Blogger Profile for the address.

Lucy Melford