It is notoriously difficult to obtain clear numbers for any sector of the spatial industry. Important shifts go unremarked simply because there is no single source where the numbers from different sources can be collated in a useful fashion. I suspect that currently we are in the midst of a silent revolution in map creation methodologies, although one would not guess as much from a casual look at the market.
I suspect that the real growth section of the market over the last few years has been the aerial survey companies, for whom business is booming. Patrick O'Connor, business development manager with Airesearch Mapping in Brisbane, says demand for aerial mapping is growing so fast that his company has put on 10 staff in the past six months. The picture from other vendors of aerial imagery is much the same. On the assumption that staff numbers can be related to turnover, this implies a growth in the market of 30% this year.
This is not at dot.com levels, of course, but it represents a healthy level of growth, especially given the expectations of spatial data managers for this year. According to a survey by Corporate GIS Consultants into the Australian and New Zealand GIS industry, data managers expect shrinking budgets next year. The Corporate GIS Survey says local government managers expect a 20% contraction, and state government managers expect to have around 10% less to spend. Those in the federal government sector are better off than most - they expect no change at all.
However, when the survey drills down a little further, to examine exactly which sectors will carry the weight of the cuts, it turns out that data managers are shifting their resources from staff and hardware expenses into data acquisition. In spite of falling budgets they will hold the line on data acquisition costs. Managers in the federal sector, in fact, expect to increase spending by a whopping 40% next year.
So, what's going on? On the face of it, it seems surprising that data managers are choosing now to concentrate their expenditure on data acquisition and upgrading. We may, however, be seeing a second generation effect occurring. That is, data managers now have working GIS to play with. Business processes are starting to depend on the GIS. Unfortunately, all too often what they are also discovering is that while the system is fine, the data on which it depends is not. Users are finding themselves embarrassed by poor quality data just when they should be milking their GIS for all the bottom line advantages it was supposed to deliver.
All of a sudden, all the old arguments about how much inaccuracy one could tolerate in the state's cadastre do not seem quite so academic. As a result, in almost all jurisdictions, there is increasing emphasis on methods of updating and upgrading existing datasets.
Against this background, a renewed emphasis on aerial photography is not so surprising. It is ideal as a rapid upgrade tool. Also, many of the traditional impediments to aerial photography are being reduced or eliminated.
Photogrammetry, once the bug-bear of any serious use of aerial photography, is now quick and cheap, thanks to new scanning technology and computer programs. This has led to a rapid increase in the amount of work around. At LandInfo, one of Australia's largest private sector survey companies, Tony Wheeler says his company spends 10% of its time on photogrammetry; worth between $1 million and $2 million per annum.
Aerial survey companies have also been quick to capitalise on the distribution strategies developed by the proponents of commercial high-resolution remote sensing from satellites. Delivery of the data by Internet or CD-ROM is now routine. Advanced ordering systems, payment-by-pixel systems, are all common.
In fact, what is happening is a complete reworking of the whole aerial survey business. Once, aerial surveying was just a way for pilots to fund a flying business. If their interpersonal skills were so bad that they couldn't handle passengers, relating to a camera seemed like a pretty good bet.
Today, however, executives in these companies see themselves as spatial data providers, who just happen to fly aeroplanes rather than total stations.
This sounds flippant, but in fact it is a key issue, because once the data is the focus of the business, all sorts of things become possible. The aim of the business now is not to get as many hours on the aircraft as possible, but to provide the maximum amount of data. Once aerial surveyors flew to order for a client. Now they fly speculatively over cities, trusting that they can sell the same data to many different clients. The result: lower costs to customers and higher profits to the surveyor.
Nevertheless, until recently, it was possible to agree that demand would increase, and costs come down, and yet still be pessimistic about aerial surveying. It was suggested that the market would be largely saturated by space-based imagery, especially when it was announced that the best of these satellites would offer sub-metre resolution.
While no one has ever suggested that spacecraft would be able to match the best aerial imagery, it was suggested that they would be so much cheaper they would drive aircraft out of the sky.
The demise of the first high-resolution supplier in Australia, the Iglass company in Brisbane, however, has shown that the assumption that satellites would be cheaper than aircraft needs to be questioned.
Iglass ceased operations 26 May, just five months after it was formed. The company was formed to market product from Space Imaging in the US. Space Imaging owns a satellite called Ikonos-2.
According to a statement from the company, Iglass had sustained a long period with no real revenues. The statement said: 'This expenditure, combined with the very weak Australian dollar, has forced the company into the position where it is no longer able to continue its operations.'
Mark Judd, a former director of Iglass, said the price of the imagery was the main reason for its slow uptake in the Australian market. He said that as a result, Ikonos data was not competitive with either aerial photography or other space-based image sources. Judd said that as a result of the collapse of Iglass, Space Imaging is currently reviewing its pricing arrangements for Ikonos imagery in the Australian market.
Paul Trezise, manager of the Australian Centre for Remote Sensing, said that the fate of Iglass did not signify the end of Australian interest in commercial high-resolution imagery. But he said the market is, and would remain, very price sensitive. Iglass had just priced itself out of the market.
The managing director of Spot's Australian subsidiary, Carl McMaster, said recently in Sydney: 'The demise of Iglass is a lesson for all of us who aspire to sell high-resolution satellite imagery.' Spot is a worldwide distributor of OrbImage high-resolution imagery, due out later this year. 'Aerial photography companies are the entrenched suppliers, and satellite imagery, to compete, has to offer advantages in terms of price and quality,' he said.
This begs the question: how much room do companies like Space Imaging have to move? It is quite true that satellites can be tasked, imagery downloaded and transmitted to the client for much less money than the equivalent operation by air. On the other hand, it costs a lot more to build a spacecraft in the first place.
How much? All the prospective companies are privately held, so it is not possible to understand their financing in any detail. But we can get some insight from an announcement from EarthWatch Inc that it had received about $300 million in new funding to see it through from 1998 to the launch of its next satellite. Suitable aircraft can be had for three orders of magnitude less.
Another point is that, for all practical purposes, aircraft operation is risk-free; at least it is low risk. There are few commercial operations quite as risky as launching space vehicles. That risk has a dollar value, as shown by the failed launch of Ikonos, and a previous attempt, by Earth Watch Inc, to put up a commercial imaging satellite.
It is by no means obvious how things will pan out in the long run. Hard numbers might be hard to come by, but one can at least put together an order of magnitude analysis of the problem. They show it will be a close thing. Roughly, if the spatial industry in Australia is worth $200 million, and that is 1% of the world total, then the world spends around $20 billion on spatial information systems. If, as the Corporate GIS survey suggests, data acquisition is about 10% of the total, then the worldwide market for data is worth about $2 billion.
In the unlikely event that raster imagery of all kinds can take 50% ofthis market then the raster market is worth $1 billion per annum. If aircraft only take 50% of this, then the portion left for satellites is worth around $500 million.
Currently, there are three other US companies poised to join Space Imaging in the market, with another six or seven in various states of development. For simplicity, if we assume there may be five operators in space within a few years, then each company would be looking at a market share worth around $100 million per annum.
Once again, being both charitable and optimistic, let us assume it costs $500 million to construct appropriate ground infrastructure, and then build, insure and fly the satellite. Let us assume a planned life of five years. One would not want much more because of the danger of technical obsolescence. It would seem, then, that the available market share would all be used up just repaying the capital cost. There is no return for the investors, and no profit.
One obvious assumption concerns the size of the market. Much of the defence of commercial remote sensing comes from a belief that once the systems are in place new applications will lead to a sudden expansion in use.
Another possibility is that there will be a very quick market shake-out. Several of the systems will go out of business, and they will either destroy their satellites, as happened to the Iridium communications network, or one operator will buy out another.
This ignores, of course, what might be the most productive market of all: the military. At least one of the prospective operators is explicitly offering a package to military operators around the world that gives them complete control of the satellite while it's over their territory - giving them the ability to scan all the borders.
If the commercial operators are able to convince the military that they are necessary, their future is assured. Even in peace-time, the world's military can be trusted to pay commercially irrational figures, especially if there is a financial threat to the data. Certainly, for middle powers like Australia, such satellites give all the advantages of satellite data without the inconvenience of having to pay the cost of one's own satellite.
This would not be a situation that would please aerial surveyors, of course. If satellite systems are underwritten by the military, they will have the ability to charge out civilian data at far below its genuine cost price. If that happens, there will b only a few winners.
