logo
Published on Media Mutandis - a NODE.London Reader (http://publication.nodel.org)

A Spatial Data Infrastructure as big as the Internet that fits in your pocket. • J.Walsh

by Jo Walsh

What is a Spatial Data Infrastructure (SDI), and why would you want one that you can contribute to and, as a user of it, control?
An SDI is made up of geographic information (geodata). Geodata can be potentially any information that describes the location of something in physical space. Geodata can be literal images of spatial things – aerial photographs taken from specially rigged aeroplanes, or satellite imagery taken from 'earth observation' satellites. Geodata can be cartographic images that are designed to look like things in physical space, emphasising certain 'features' that are most useful to the intended user of the map – streetmaps for cities, or contour maps for exploring the countryside. Geodata can be your location as measured by the Global Positioning System (GPS); geodata can be your address; geodata can be any kind of information that has a “spatial component”.

A frequently repeated factoid is that 80 percent of information collected by government agencies has a spatial component. Governments need to collect a lot of geodata in order to do the work they currently do – collecting taxes on property, managing zoning and planning of land developments, managing their electoral registers, telling people where to go and vote, deciding whether people qualify for health and welfare services, and if so, how they can best go about getting them. Governments at every level need to collect and analyse a lot of geodata. Geodata even describes how government works, by drawing invisible political boundaries that determine how different people “represent” you in government, in a local council, or a city council or members of a national assembly of representatives.

Most countries have a “National Mapping Agency” (NMA) that collects and manages geodata on behalf of different government departments. The UK's Ordnance Survey is probably the world's oldest National Mapping Agency. The name “Ordnance” offers an insight into its history; it was set up to create maps for military defence purposes. Maps are the most powerful military technology there is. Now there is a new set of technologies for making digital maps, NMAs have decided they need a National, or even a European or Global, SDI for sharing geodata between different people and organisations.

Over the last 25 years, mapping techniques have been changed by computers unrecognisably. Satellites orbiting the Earth take pictures that can be as accurate as 1 metre per pixel – a metre of distance on Earth's surface corresponds to one little pixel dot on your screen. It's not hard to imagine a digital map of Earth that is the same size as it, stored on a computer network, visible one small piece at a time. Other networks of satellites let you find your position on Earth's surface to around 5 metres of accuracy with a GPS device. There are other ways to find your position by listening to signals from different Earthbound radio transmitters whose locations you know. With a location sensing technology connected to a wireless communications device – like a mobile phone or a laptop – you can find a lot of instant information about where you are.

Just looking at a picture of where you are isn't very interesting: you can do that with your own eyes! Connecting different kinds of geodata together, you can find out a lot more about your environment than you can see for yourself or learn from people nearby. Mobile phone companies hope to sell you “Location Based Services” that tell you about shops and restaurants nearby. They also hope to sell “e-government” services to people responsible for governing a space, though no-one is sure what e-government actually is yet.

There is so much more potential in what geodata can help you find out about your surroundings; images of what the area looked like 50 years ago, stories written by people who used to live there, how different people have used the space you are in, made plans for the space and seen those plans overwritten by the passage of time. Now as never before in history, every person has the ability (if they can afford to buy or make the right bits of technology) to explore and contribute to a “collaborative map” showing how people and things interact in space over time.

In order to draw a collaborative map together, one person needs to agree with another that they are talking about the same thing. We can decide to give labels to things – the strip of tarmac outside this house is labelled “Massachusetts Avenue” – but when we talk to a third person, we need an easy way to teach them what the labels are and what the labels mean.

People trying to create knowledge together on the World Wide Web needed to define “standards” which say, more or less: “when we use this set of labels for things, we agree that we are all talking about the same thing”. HTML – the HyperText Markup Language – is a simple standard for talking about different parts of documents – what a title is, what a paragraph is – in a way that makes it easy for computers to show humans the “same thing” when they look at a document. In order to keep the standards honest and make sure that they work in the same way for as many people as possible, “standards bodies” come together to represent different views. Standards bodies are a kind of government system for deciding how information is transmitted from one place to another.

As geodata is becoming more interesting and important, more people want to create standards that make it easier to share. One important standards body is the “Open Geospatial Consortium”, a group of companies making software that helps people work with geographic information. They publish a Geographic Markup Language which can be used to describe spatial things. Arguably the most important standards body on the web is the World Wide Web Consortium, or W3C, created by Tim Berners-Lee, whose work on standards made the web possible in the first place. His vision is of a “semantic web”, which allows people to more easily and reliably create and exchange meanings of labels for things with software.

Wikipedia has an nice definition of what an “infrastructure” is; a kind of internal framework that allows people to transport knowledge, or things, or other people, from place to place. Standards for describing information are the bricks out of which an “information infrastructure” is built; we would not be able to share definitions and decisions without them. Standards for describing geodata – where things are and how things relate to other things in space – are the basis of what is called a Spatial Data Infrastructure.

Is an SDI really necessary to share data? Different governments view this in different ways. The US government gives away a lot of the geodata it collects through its Geographic Survey and Census Bureau, allowing anyone to create their own maps and “locative media” with it. The UK's Ordnance Survey, like most European National Mapping Agencies, sells temporary rights to use its geodata at the highest price it can get away with. So in Europe, a lot of amateur programmers and enthusiasts have come together to build shared, open maps from different sources of geodata – “tracks” that show where a GPS unit has been, low-quality satellite photographs published by the US government.

The Open Geodata Manifesto talks about the social and economic reasons why open access to geodata should be considered a right, not a privilege. It lives at http://okfn.ogr/geo/manifesto.php [1]. It offers these suggestions as the core of an open geodata policy:

All government-collected geodata should be open, that is, available for free distribution and re-use under a ShareAlike license.

Online mapping projects creating freely reusable geodata should offer a compatible open license.
Common, standard formats for describing and exchanging geodata should be adopted.

Ultimately, all state-collected information should be openly available, in a structured machine-readable format.

Because there weren't standards really designed for collecting data from a lot of different, potentially unreliable sources, Openstreetmap had to create its own. As the industry standards bodies wake up to the new things that people want to do with maps and geodata, their own standards are improving, changing in ways that make it easier for people to write software to share geodata between each other.

So perhaps it's easier, faster, more fun and even more “correct” to put geodata on the Web just as we put documents on the web now, and create a “geospatial web” together, rather than letting governments try to build expensive SDIs that are designed to “protect” geodata from people rather than let them share in its creation. Chris Holmes neatly summed up the possibilities here at the end of the Open Geodata stream at Wsfii:

"Geodata should be something that is taken for granted as a base for further free information infrastructures: open source traffic modelling, open source environmentalism, natural open source resource management. The geospatial web should be a texture, just like the internet is the basis of many other infrastructures on top of it.

We're only going to only get there together, we need to build this, we need to free the spatial data infrastructures from the powers that be, tailor them to our uses and build something better than they could even possibly imagine".

first draft 19-01-2006 - second draft 23-01-2006
 


Source URL:
http://publication.nodel.org/A-Spatial-Data-Infrastructure