¿ìè¶ÌÊÓƵ

How much is a map worth? Within hours of the flaming debris of Pan Am
flight 103 crashing onto the Scottish village of Lockerbie on 21 December
1988, the emergency services had large-scale maps of the area and its surroundings
on the scene. The maps, which came from the Ordnance Survey, Britain’s national
mapping service, were essential for coordinating search teams in the remote
Scottish countryside and, later, for plotting the precise position of every
casualty and piece of wreckage. This information proved vital in tracing
the cause of the disaster to a bomb in a cargo hold.

The effectiveness of those detailed maps highlights the principal issue
facing the Ordnance Survey (OS), following the return to power in April
of the Conservative Party with an electoral mandate to pursue its programme
of privatisation. Should the government continue to fund the organisation
responsible by law for mapping Britain?

David Rhind, the head of the OS, has some answers. Among the papers
that awaited Michael Howard when he took over as environment secretary was
a report from Rhind looking at various ways of financing the OS. Rhind
had been asked by Howard’s predecessor to explain how the OS could cover
all its costs – a prerequisite to making the organisation an attractive
buy should the government decide to sell it. In the past, he has criticised
both the OS and the financial support it has had from the government. But
after barely six months in the hot seat, he appears to see a good case for
funding after all.

‘Take Lockerbie. That’s a semirural area; not many customers. If I had
said, weeks before the plane crash, that I was going to stop producing
that map unless people paid for it, I don’t think many people would have
come forward,’ he says. But coordinating a search in darkness with out-of-date
maps is no one’s idea of the efficiency that privatisation is meant to bring.

CIVIL RECORD

Though the organisation has recently acquired some autonomy, its headquarters
in Southampton retain the look of the civil service, from the building’s
1969 concrete design with its fluorescent-lit corridors to the notice announcing
‘staff shortages’ that periodically close the map shop in reception. If
that sounds musty, the new head is definitely not. Rhind is the first professional
geographer to lead the 2400-strong staff. His predecessors were land surveyors,
and until 1977 came from the military. Before joining the OS, he was professor
of geography at Birkbeck College, London, with extensive experience in the
application of computers to his subject, especially in developing geographical
information systems.

A GIS is a system that manipulates spatial data to solve mapping problems.
It creates information about a location by overlaying sets of computer data
(such as topography, roads, addresses, soil quality and rainfall) for that
area. For example, architects planning a new supermarket can choose the
best spot by overlaying data sets of transport access and the demographic
breakdown of nearby housing estates onto a basic, digital map. The benefits
of a GIS often depend on its user’s imagination: as Rhind points out, just
20 data sets can be combined in more than a million ways although, as he
notes, ‘you probably wouldn’t find it much use combining a data set of geology
with unemployment’.

From 1984 to 1987 Rhind was a member of the Chorley committee on GISs,
set up by the government to examine the potential for computerising data
on land use and ownership. He also served on and advised other government
committees looking at developments in GISs, and the future of the OS.

One of his favourite phrases is ‘the past is not the key to the future’.
He applies it often to the work of the OS. Though his office contains an
ageing globe and a bookcase full of the familiar, if now outdated, pink-backed
maps covering the whole of Britain, he sees these treasures as old technology.
New technology is changing map making radically. In 1980 the OS completed
its 35-year postwar remapping of Britain; the development in that time
of satellites belonging to the Global Positioning System, which can locate
ground objects to within a metre or two, means that this enormous task need
never be repeated, although smaller changes will still have to be noted
by the OS’s 750-odd field workers on the ground. Within a year, Rhind says,
they will be taking notebook computers out with them, to update maps on
the spot. He envisages ramblers of the future determining their exact location
from hand-held computers. It might take the fun out of orienteering competitions,
but that’s progress.

Yet the biggest change of all, and the one whose effects will shape
the OS’s future, came in 1971 when the organisation began moving to digital
mapping.

The task of converting all the information on every map into digital
code was enormous, and Rhind points out what a brave decision it was. ‘Making
maps digitally then was far more expensive than by hand. Even in 1979,
it was 30 or 40 per cent dearer.’ Now, the benefits are clear: updating
the database costs about half as much as redrafting on paper, and as the
amount of data logged rises and the cost of labour goes up, that gap will
increase. By 1995, the £25-million project to digitise the data from
220 000 differently scaled maps of Britain – Rhind’s predecessor, Peter
McMaster, boasted last year that ‘every back garden, shed, letter box and
telephone kiosk is on them’ – will be complete.

Digitised maps are surprisingly undemanding of computer space. All the
back gardens and letter boxes in Britain should fit into about 100 gigabytes
(100 000 megabytes), which is within the capacity of most mainframe computers.
A map at a scale of 1:50 000 typically requires 20 megabytes, the average
storage capacity of a personal computer. Rhind boasts that nowhere else
has such detailed coverage reached the same stage of completeness in digitising.
He cites the example of the US, where the digital map of the country completed
by the Bureau of Census only covers elements at a scale of 1:100 000. At
the Ordnance Survey, the figure being aimed at is 1:1250, making it more
than 6000 times as detailed. The US has only completed 10 per cent of the
work for more detailed digitisation, while the OS has done 80 per cent.

MAP MARKETING

‘The role of the OS now is to maintain the information, rather than
create it,’ Rhind says. ‘All the OS’s future is about spin-offs from the
database.’ This will make marketing more important than ever, as is already
becoming clear from the changing character of the organisation’s staff.
While employee numbers have fallen by a third since 1979, as fewer field
workers and map drafters have been required, the marketing department has
grown from around 80 to 167.

Does that future also include privatisation? In a paper presented in
May last year to the Survey and Mapping Alliance, an organisation that brings
together representatives of government, industry and academe, Rhind turned
his critical academic eye on the future of the OS and publicly questioned
the need for the government to fund it. Within four months Michael Heseltine,
the then environment secretary, announced Rhind’s appointment to the organisation’s
top job. If this makes him look like a political appointee, Rhind says he
had deeper motives for accepting a job he had liked the look of for some
years. ‘There’s only one like this worldwide, and only perhaps half-a-dozen
comparable organisations.’

He didn’t expect a warm welcome. ‘I came with a little bit of hesitancy,’
he says. ‘I’ve been involved with the OS for 20 years, sometimes as a consultant,
sometimes a critic of what I perceived then as traditionalism and overconcern
with the paper map. Their argument was that no one was buying digital, but
people were buying paper maps.’

But he was pleasantly surprised. ‘I’ve fallen with my bum in butter,’
he says. (That is, he’s happy.) ‘The people here are very outward looking;
some make me seem very fuddy-duddy. They’ve got a lot of ideas about the
future. I had thought I might have to come as a critic and a single, lone
voice – I found I’m among people who understand.’

It seems a very dramatic conversion, but Rhind denies suggestions that
he is just covering his back from any in-house knives, or trying to present
a united front for his new troops. ‘On arriving, I found a lot more revolution
going on than had ever been apparent. The OS was very careful about what
people outside knew, but things were more advanced inside.’

So was his past criticism of the organisation based on flawed perceptions?
‘No. It was fair, based on the perception the OS gave and the talks that
had been given. But inside, there were people working to make the right
things happen.’

They have a long history behind them. The OS was set up in 1791, primarily
for the army, which wanted detailed knowledge of the coastline and positions
to be fortified against a threatened invasion from France. The first OS
map, of the county of Kent, was published on 1 January 1801, at a scale
of 1 mile to 1 inch (1:63 360). The OS remained a military stronghold for
nearly 200 years. Its first civilian director-general, Walter Smith, took
office in 1977 and when the last military personnel left, in 1983, the army
literally beat the retreat.

HIDDEN TREASURE

That year the OS was told by government it should generate more revenue
through marketing its expertise. Joint ventures followed at a growing pace,
from birdwatchers’ guides to motoring atlases. In May 1990, the government
made the OS a semiautonomous agency, and told it to recover 65 per cent
of its £70 million annual budget. A year later, the ratio was increased
to 70 per cent. Rhind sees strong commercial logic in this and derides mapping
authorities that fail to exploit their assets. For example, US law states
that because the Bureau of Census is funded by federal taxes, the price
of the maps it sells must not exceed the cost of reproduction. Thus anyone
can buy its complete digital map of the US for $1500 ( £790) on magnetic
tape, or $32 on CD-ROM. The private sector can make enormous profits simply
by repackaging the information, says Rhind. ‘In Britain, any information
is seen as a commodity. Selling it is a way to lighten the burden on the
taxpayer.’ Though the OS has not yet fixed a price for a fully digitised
map of Britain, its pricing policy is geared to achieving the government’s
rising target for recovering costs. In comparison, the US Bureau of Census
recovers 2 per cent of its cost while France’s Institut Geographique National
(IGN), its equivalent of the OS, manages 39 per cent, he says.

Rhind points to the drawbacks in not putting a price on information
by contrasting the accident at Lockerbie (for which the maps, not yet digitised,
are stored on microfilm) with the 1989 earthquake in San Francisco, which
tore apart the coastal highway. In California, the available maps were not
detailed enough for the emergency services to use. Instead they had to
rely on more detailed ones produced by local map makers, which were inconsistent
at their border points. To Rhind, that is clear evidence that the US system
doesn’t work: the mappers make too little profit to invest in the wide-scale
schemes vital in emergencies. He points to the advantages in making the
Post Office, gas, electricity and water undertakings pay fully for the maps
they use. ‘I don’t want it to sound rude, but I think it’s reasonable that
the utilities should have to pay for some things.’

This leads back to the central question of a map’s value. Rhind reckons
it would cost ‘hundreds of millions of pounds’ to replace all the OS’s maps.
But that is not their true worth: ‘Value comes from what you make of it,’
he says. As digital maps become more widely used, the OS will profit more
by exploiting its copyright on the data, rather than on only the maps it
sells.

As the delivery of a fine-detail digital map of Britain draws closer,
possible applications for it are multiplying. For example, British Telecom
has half a million paper maps and diagrams of exchanges, cables and subscribers,
and employs 2000 staff to keep them up-to-date. If BT wants to go wholesale
to a digital format, the OS can almost name its price: once overlaid on
the basic map, the diagrams of the network will need little updating. The
same applies to the other utilities, as well as emergency, security and
delivery services. Similarly, architects and engineers want to put all the
information about a particular site as well as their design work on one
computer.

Even ramblers are catered for. Already they can walk into a shop in
central London, the National Map Centre in Westminster, and buy a map printed
out on the spot, to any scale (because scale is infinitely variable on digital
maps, down to the basic level of data originally collected) centred on the
particular location they want. Later this year, the OS will have three such
agents; by next year, 15. The shops will lease the equipment – a computer
workstation, CD-ROM reader and colour plotter. For the OS, it is better
than guessing how many printed maps of which areas will be needed.

But competition is growing. Because it receives public money, the OS
cannot refuse to sell its raw data to commercial customers, who may then
be able to find better applications. Laser-Scan of Cambridge, which provides
a map-digitising bureau and expertise, has developed a GIS based on OS data
that is used by the University of Stirling to analyse the environmental
effects of proposed power lines. Nextbase, a software company in Ashford,
Middlesex, uses a compressed version of the OS’s 1:625 000 digital map of
Britain as the basis of its computerised route-mapping system, Autoroute.
The program runs on a standard personal computer, with the geographical
data squeezed into 800 kilobytes. Users enter the start and end of their
journey, and the program works out the shortest or theoretically quickest
route. Meanwhile, more and more competitors are producing small-scale road
maps. Rhind also finds contenders among companies producing products at
larger, more detailed scales – ‘but only where there’s a commercial prospect’.

For the OS, all this creates problems born of ideology rather than ability.
If it is privatised with a remit to keep surveying and producing national
maps, regardless of exact demand, it will have to charge higher prices to
cover the costs of mapping the non-commercial areas that rivals do not bother
with. But the rivals that use its data could then complain that the OS
is exploiting its position, and not charging in line with the cost of producing
the data.

Rhind says the answer must lie with the government: is it prepared to
pay for the uneconomic aspects of national mapping? ‘The more competition
we have, the more I have to react in the same way – which potentially means
closing down mapping of remote areas. Competition is more significant than
government ministers have realised until recently.’ Any concern that the
OS has a monopoly on national map making should, he says, be tempered by
the fact that there’s ‘only a monopoly in areas where there’s no money to
be made’.

Meanwhile, foreign rivals are multiplying too, including satellite
mappers such as Landsat of the US and France’s SPOT, which can create stereoscopic
pictures to provide three-dimensional images. They are ‘reasonable for mapping
updates at scales of 1:100 000 and maybe 1:50 000’, Rhind admits, but he
dismisses the idea that this development is a commercial threat. ‘For the
next two or three years Landsat and SPOT are not directly relevant to the
OS’s activities,’ because the organisation is more interested in selling
detailed or digital maps. And after that? When pushed, he admits that the
OS’s research and development teams are looking at generating three-dimensional
models from the SPOT stereo pictures. But he still sees the best application
of such techniques being outside Britain, where they might be used by the
OS’s overseas division, which produces maps for foreign clients. Three-dimensional
mapping only becomes viable if an area has not been mapped already.

OVERSEAS EXPANSION

He also thinks the overseas division is a clear candidate for expansion.
For example, the OS recently completed a project to create a land registry
for Bulgaria, recording the position and ownership of every plot of land
in the country. But this leads him into a position on subsidies quite unlike
the academic one he advocated in May 1991.

The major obstacle to expanding the overseas division is that better-funded,
state-run rivals can undercut the OS because their governments see an extension
of their mapping business as a matter of foreign policy. ‘It’s a pretty
ruthless game out there,’ Rhind reflects. The OS receives funds from the
Overseas Development Administration, but Rhind still feels that government
support gives rivals from France and Sweden an advantage. France’s IGN has
even tendered for aerial mapping work over Scotland. ‘Well, we can’t be
Fortress Britain,’ he sighs.

Another way to keep the OS growing in the face of such competition is
merger. One possible candidate is the Land Registry, because it relies
heavily on the OS’s data and the Conservative manifesto cloudily mentioned
‘exploring the idea’ of a new computerised property data bank, combining
information held by the Land Registry and ‘other bodies’. In short: merge
the registry and the OS, with the OS as the senior partner. Rhind remains
detached, but would not be surprised to see civil service infighting against
the idea from the Lord Chancellor’s Department, which runs the registry.

Still, amid the ideological battles, he notes that ‘what brings it all
together is geography’. You sense his relief. ‘In any database, you have
to have a common key – whether it’s a grid reference, or the latitude and
longitude.’ And the OS rules supreme in British maps. Rhind insists that
is not because he wants to run a cartel, but because it is essential that
a single organisation is in charge of such an important commodity.

To emphasise his point he recalls an episode when he helped build a
‘green’ database for the European Commission. The aim was to record soil
productivity across Europe, using data sets with maps from different countries.
It was a fine idea. But there were hidden mismatches in the borders of the
digital maps. When the proud moment came and the data sets were overlaid,
the onlookers discovered that some of the soil was apparently located in
the sea.