Abstract
Location Based Services (LBS) will have a major impact on the cartographic community because of the increased demands for map accuracy. For automotive applications (route finding and in-car navigation), 50 metre accuracy has become the norm. However, for pedestrian use - especially in an urban environment - spatial accuracy in the order of 5 metres is / will be required. In addition, there is a growing demand for "content" which is the word now being used to describe all the non-geographical detail associated with location including house numbers, details of one-way streets, postal districts, rising bollards and similar. The collection, representation and maintenance of both the spatial and non-spatial components of data for LBS applications are becoming a major new activities for producers of geographic data for the new economy. Also, the cartographic industry has an important role to play in defining the standards for data collection, data capture, data storage and representation.

Although cartography has its roots in the the production and distribution of paper maps, it is increasingly embracing the digital revolution. In some respects, this metamorphism mirrors what has happened within the various National Mapping Agencies - the NMAs.

Initially, the NMAs utilised digital techniques for improving the production flowlines for paper maps: this then evolved into the distribution of mapping data in digital form for use with one or another of the GIS systems for asset management, logistics planning and a myriad of commercial applications.

However, another revolution is now upon us which takes the delivery of digital maps directly to the end user. We have already seen evidence of this on the Web: for example, MapQuest, Maporama, MultiMap and Michelin - just to focus on the "Ms" - all provide mapping over the web, together with a variety of ancillary services e.g. for route finding, housing information, local garages and similar.

However, this is all going to change in the very near future as maps become increasingly available over mobile telephones.

The mobile revolution will have a major impact on the requirements for map data for a number of reasons - but, mainly to do with the fact that the majority of telephone calls are made by pedestrians rather than by true mobile (i.e. in-car) users. Some of the more significant of these new requirements relate to map accuracy and to "content".

Map accuracy
Currently, in-car use of mapping demands positional accuracy of approximately 50 metres. Automotive requirements will become more demanding as in-car navigation evolves to include lane-keeping and collision avoidance technologies. However, pedestrian requirements are an order of magnitude more stringent.

Consider just two scenarios: a simple (pedestrian) route finding application must take account of which side of the road you are standing on in relation to e.g. a pedestrian underpass. Even more important, a hide-and-seek game must "recognise" which doorway your other game players are using to avoid capture.

In case anybody is any doubt - these are applications which are being deployed right at this minute. The big revenue earners for the future mobile radio industry include sex, shopping and soccer (sometimes described as girls, games and goals). Each of these, in its own way, relies on highly accurate maps delivered over a mobile radio network.

Pedestrian-based applications of mapping will demand sub-five metre accuracy in the urban environment to satisfy the market for Location Based Services.

There is, however, a further component of "accuracy" which is related to customers in a pedestrian environment. This is the extra detail which will be required: the need to include small alleyways (short-cuts), the internal layout of parks and gardens, and the various pedestrian subway systems. Subways are particularly important for multi-modal transport.

For example, a route-finding application may produce a route which is a combination of bus and underground train, followed by a walk through an old town cris-crossed by small passageways (numerous examples come to mind in e.g. Krakow, Brugges, Budapest and Heidleberg - to mention just four historic towns with a long history of catography. In a situation like this, the amount of detail which is required far exceeds that which is available in a conventional (car) navigation system.

Fortunately, the advance of technology, means that it is now possible to satisfy demands for enhanced accuracy at a relatively modest cost. As part of the Galileo programme, ESA has recently deployed a pilot version of EGNOS in Europe (WAAS - or Wide Area Augmentation System - in the USA): this system is, in many ways, similar to DGPS (Digitally Enhanced GPS) and provides sub 2-metre accuracy across the whole of Europe. In fact, Geo Strategies provided the Eastern European benchmark information at the first Galileo conference in Warsaw (May 2003) which demonstrated the commercial benefits of using EGNOS for a major re-survey programme across those countries bordering the Black Sea - namely Bulgaria, Romania and Moldova. This will be extended to Hungary, Slovakia and Slovenia in the near future.

Content
A new word - "content" - is increasingly used by data producers which work with the web mapping market and, especially, with those of us supplying mapping data to the telephone industry.

Content is generally taken to mean all that non-geographical detail which is associated with location. This can take many forms such as road numbers, details of one-way streets, the presence of rising bollards, postal districts and census information. In addition, there is a whole class of content called "Points of Interest" or POIs.

The range of Points of Interest can be huge. In our databases for Central & Eastern Europe we have 136 classes of POI, ranging from embassies, hotels, restaurants and bars, to cemeteries, monasteries and vineyards. However, this is a very modest number in comparison with lists of POIs for the more developed economies.

Each POI has a name, a location (geographical coordinates) and attributes such as opening times, telephone numbers, types of food (for restaurants), etc.

The major problem with content is that, again, the main users will be pedestrians and that means that their location on a map (the point to which they are geo-referenced) needs to have similar levels of accuracy i.e. in the order of five metres.

Thus, the collection and maintenance of content has become a major activity for producers of geographic data for the new economy. Simplistically, it is based on the geo-referencing of e.g. Yellow Pages but, in practice, it goes far beyond that. The most obvious differences relate to the extended attributes that are now being required and also the non-addressable POIs - non-addressable in the sense that you never send a letter to them e.g. statues, monuments and outlook-points - objects of interest that have a location, but no address.

Data structure
There is a further consideration that must be taken into account for geographic data to be used for LBS - the data structure.

Consumer-based applications (location-based games and similar) absolutely must match consumer expectations: in practice this means that they must look good and be displayed quickly. People have become increasingly demanding in terms of the look, feel and speed of delivery of maps over the web: this expectation becomes raised to least to the power two - if not three - for maps delivered over a mobile phone.

In turn, this has resulted in the creation of very specialised data formats which are efficient for phone-based delivery based on the principle of a "thick server" and a "thin client" i.e. with all the processing being carried out (very quickly) by the service provider, and the display being carried out simply and efficiently in the handset.

These formats (e.g. GDF or ISO TC204) are very precisely defined and are progressively becoming the standard exchange format for Location Based Services.

Summary
Location Based Services have been rather slow to take off in the mobile telecom environment: they have been caught up in the general economic downturn of the telecom sector. However, they represent real opportunities for cartographers in terms of the improvement of both map detail and map accuracy, plus the enormous task of capturing and maintaining content databases which will become the real differentiator between service providers.

We have to accept that high precision mapping is fast becoming a commodity and that the value added components represent new opportunities for the cartographic industry.

For further information, please contact Geo Strategies at:
	Geo Strategies Ltd			Geo Strategies SA
	St John's Innovation Centre			Str G-ral V. Milea, 10A
	Cowley Road			Sibiu 550331
	Cambridge  CB4 0WS			Romania
	United Kingdom
	Tel:	+44 (0)1223 205080		Tel:	+40 (0)269 210832
	Fax:	+44 (0)1223 205081		Fax:	+40 (0)269 211165
	Email:	maps@geo-strategies.com		Email:	maps@geo.strategies.ro