The system, known as the Australian Asian Visible Imaging Spectrometer (AAVIS), is essentially an electronic eye or scanner which can capture digital images of the Earth’s surface in fine detail. It is able to analyse an area of land or sea as small as 12 square metres, compared with a minimum of 100 to 900 square metres for satellites.

The electronic eye – which is mounted on an aircraft – can distinguish 40 to 70 times the number of individual colours in an image than can be achieved by instruments carried on satellites, according to Steven Chadd from the Melbourne engineering firm, Trippett Shedden. This capability will allow users, for example, to distinguish easily between toxic blue-green algae and benign green algae within a single lake or river.

“Our goal is to reach a level of spatial and spectral resolution which will indicate the health of whatever is being looked at,” Chadd said.

The power of the instrument is due in large part to scientific algorithms developed by Australia’s chief research agency, the CSIRO, in collaboration with Trippett Shedden, and the Australian Institute of Marine Science in Townsville. Indonesia, which is seeking to boost its expertise in advanced electronics for the aviation industry, is collaborating in the development of a commercial product which can be marketed throughout the Asia-Pacific region (see Box 2).

Research over the past decade, principally by CSIRO’s Institute of Natural Resources and Environment, has culminated in the development of image- processing software which can extract detailed information from an image such as the condition of plants and other organisms growing in a paddock, or even on the bottom of a lake.

Organisms reflect light in a particular way. This characteristic, called a spectral signature, allows them to be identified. In visible light, these signatures would be called colours or fine shades of colour, but they also include parts of the infrared spectrum invisible to the human eye. CSIRO’s understanding of the spectral signatures, combined with the physics of how reflected light travels through and is absorbed by water, allows scientists to see into a body of water from the air, and determine the presence or absence of particular organisms at a depth of up to 15 metres.

Using a prototype of the AAVIS system, Trippett Shedden recently surveyed Melbourne’s Port Phillip Bay to map areas where scallops may live and regions containing seagrass mats, which provide a home for many species of commercial fish.

“From the air, in a period of about five days, we surveyed the entire bottom of the bay except for the deep channels,” said Chadd, who is Trippett Shedden’s project manager for AAVIS. “We can tell you that the water is the colour it is due to a particular organism living on the bottom.”

The prototype system uses a scanner built by a Canadian company, Itris Research. The scanner is able to distinguish 288 spectral bands or “colours” of light in the visible and near-infrared electromagnetic spectrum. A more powerful instrument is likely to be built in Australia, and will be able to distinguish 512 spectral bands. By comparison, the widely-used Landsat satellite imagery contains only 7 spectral bands.

Chadd claims AAVIS has three main advantages over satellites: more spectral data is possible, it can be used when and where needed, and the data produced is owned by those using the system, whereas satellite is available to anyone willing to pay.

AAVIS also promises quicker results than surveys done on the ground or by boat. A survey of algae in Sydney’s Hawkesbury River was completed by Trippett Shedden in three days – a task that would normally take divers months or even years to complete to the same level of accuracy.

The company plans to offer a commercial remote sensing service internationally within the next two years. The service will provide information such as whether a coral reef is living or dead, whether a wheat crop is healthy or stressed by drought, or whether a grassland is becoming environmentally degraded.

AAVIS is one of a number of aerospace technologies, including a network of ground stations for satellites and a service to predict changes in the ionosphere, which are being considered for possible joint development between Australia and Indonesia as part of an agreement known as COSTAI (see page 5).

Box 1

A SIGN that the Indonesians were particularly interested in AAVIS came last month when Trippett Shedden, an engineering firm in Melbourne, was invited by Indonesia’s Agency for that Assessment and Application of Technology (BPPT) to take part in the initial planning of a scheme combining industry, agriculture and hydroelectricity generation in the remote Mamberamo Valley in northwestern Irian Jaya.

But the scheme could run into problems. In the past, conservation groups have claimed such schemes were not suitable for tropical regions. The creation of large bodies of water in warm regions can cause the sudden influx of new species, particularly birds and insects. These can have a detrimental effect on indigenous species and may introduce disease.

Indonesian officials are likely to argue that, by using AAVIS, the scheme can be designed and managed in an environmentally sensitive way.

But Australia is giving tacit endorsement to the scheme by assisting with the introduction of AAVIS. “If Indonesia is going to get into a massive development project, then it would be sensible for them to do some surveying and that is where we can help,” says David McEwan, of the Department of Industry, Science and Technology’s International Collaboration Branch. In its commercial dealings with Indonesia, Australia is not in a position to make a moral judgment on the way the country develops, he said. “We can help to repair the effects of rapid development that has occurred and help plan development to avoid problems in the future.”

Box 2

LATE LAST year the deputy chairman of Indonesia’s Bureau of Industry Strategy, Dr Wisnubroto, said: “Indonesia has an aerospace industry but no infrastructure. Australia has an infrastructure to support a global aerospace industry. So the two should be able to cooperate for mutual benefit.”

It is this thinking which lies behind the joint venture to develop a high resolution imaging technology.

Over the past 18 years Indonesia has poured about A$1.5 billion into the creation of an aviation industry. But it is struggling to compete on the world market because of a lack of support industries, such as advanced electronics, and a dearth of skilled workers.

Just across the Timor Sea, Australia performs world class research and development and has well established training facilities. But it only has a small manufacturing base and a limited local market.

Bucharuddin Jusuf Habibie, Indonesia’s Minister for Research and Technology and the driving force behind the aviation industry, wants Indonesia to be a collaborator in high technology ventures, not just a consumer. The AAVIS project is appealing to Indonesia because it could lead to a commercial system that could fly on an Indonesian aircraft. The aim will be to market a service based on AIVIS throughout the region.

Until now the Indonesian education department has only given competitive research funding on a one-year basis. The new 16-member council will finance longer term projects. It will be funded initially by a World Bank loan of A$80 million for five years. The Indonesian government is planning to add another A$55 million. Indonesia has only 22 institutions offering postgraduate training and able to receive the URC funds.

The delegation – led by the deputy chair of the URC, Jayah Koswara, who has served on higher education boards and panels within the Indonesian Department of Education and Culture – spoke to Australian Research Council chairman Max Brennan.

The DSTO – set up in 1974 to undertake defence-related research that was
previously administered by four government departments – has an annual budget
of about A$200 million and a staff of 2900. This includes 1200
scientists and engineers who work mostly in laboratories in Melbourne and
Adelaide.

The organisation wants to lift its revenue from commercial ventures tenfold
over the coming decade, from around A$2 million a year to A$20
million. It will not be easy. The DSTO, like the CSIRO and the universities,
is under pressure to boost commercial income. But to date the results have
been mediocre with the A$2 million from royalties, licences and other
agreements barely covering the expenses. According to Ian Hagan, head of the
new Commercial Activities Division, 1994 will be the first year that the DSTO
will start to show a profit from its commercial ventures.

Some of the hurdles faced by DSTO stem from the rules that have been put in
place by successive governments more concerned about Australia’s defence needs
than making products from DSTO inventions.

All major decisions, such as changes to the relationship DSTO has with a
company, need to be approved by no less than three ministers – the ministers
for defence and finance and the Attorney-General. “This means we are not
always operating in a timeframe it takes to make a commercial decision,” says
Hagan.

But impediments to commercialisation are being removed. Until recently, the
DSTO could not spend more than one per cent of its budget trying to generate
commercial revenue. This restriction, according to Hagan, gave industry the
impression that DSTO’s commercial activities were a poor second to serving the
country’s defence requirements. In future, the level of investment is to be
based on what makes business sense, not on a finite figure.

Hagan’s job will be to present the DSTO’s case to government for changes to its
commercial operating procedures and to coordinate the broader range of
interactions with industry. His division will take control of the Industry
Support Office. The office was set up last year and is headed by a marketing
consultant from outside the DSTO – Roger Bouette from Invetech in Melbourne.
It is meant to be the first port of call for DSTO scientists wanting to reach
industry.

As part of its new direction, the DSTO is shedding non-scientific staff to
concentrate on research activities. For the last few years, it has lost
support staff at rates of between 100 and 200 a year. This will continue for
at least another 18 months.

The DSTO refers to what it wants to do as “conceptual research”. It will take
research to the prototype stage but then farm out the engineering to industry
under a joint development arrangement. Instead of handing over the rights to
an invention and then forgetting about it, the DSTO says it wants to work
closely with companies to refine the technology and develop a product. The
black box flight recorder and photocopying are two DSTO technologies that have
slipped away from Australian industry in the past.

The development of the Jindalee over-the-horizon radar is an example of the
new approach. The DSTO produced the concept and Telecom Australia took over
the development. Similarly, DSTO research on high-grade steel, in
collaboration with BHP and Bisalloy, was used in the construction in Adelaide
of the Collins Class submarine. The Collins is the fastest and most
technically advanced non-nuclear submarine afloat. The development of
materials used in the Collins is expected to lead to a new range of exports.

Pat Clark and Warren Marwood of the Defence Science and Technology Organisation
in Salisbury, near Adelaide, say that their invention could offer a cheap
alternative to buying time on supercomputers. It could, for example, bring
real-time medical tomography into GPs’ surgeries and local health clinics,
they say.

At the heart of Clark and Marwood’s invention is a chip they call a
scalable processing array, or SCAP, which contains a rectangular, interconnected
grid of 20 processing elements. SCAP systems that have been built to test
the concept contain 20 of the SCAP chips arranged in four rows and five
columns, giving a total of 400 processing elements.

The matrix design provides a fast way of manipulating arrays of numbers
or data arriving in a parallel stream. In conventional, linear processors
such data must be broken down before processing. The researchers say the
only limit to the speed at which data can be processed by a SCAP array is
how fast it can be fed into the edge.

Clark and Marwood say that around 80 per cent of calculations now performed
on supercomputers use matrix algorithms, a technique that can deal with
extremely complex mathematical problems. ‘The advantage of reducing problems
to matrix form is that it controls complexity,’ says Marwood. ‘Matrix algebra
can also provide results that are not obtainable in any other way.’ The
SCAP design provides a very fast way of performing matrix algebra because
whole rows or columns of numbers can be processed at once.

Development work on SCAP began in the early 1980s, aiming to give the
military a compact means of processing massive streams of data such as the
sonar echoes received by submarines. The rapid coding and decoding of security
information is another military application. But Clark and Marwood suggest
that SCAP is also well suited to civilian applications such as computer
graphics and virtual reality simulations, which can require enormous processing
power.

The Adelaide company Radlogic has worked with Clark and Marwood in the
design and construction of the SCAP hardware. Robert Clarke, Radlogic’s
managing director, says that plugging in a SCAP can boost the performance
of a Sun workstation by a factor of between 10 and 100, to 150 million floating
point operations per second. An Intel i860 chip, normally used for image
processing, provides 80 million floating point operations per second.

But according to Clarke, ‘the real advantage of SCAP is that if you
want to expand the capacity to make the system go faster you simply expand
your array by adding chips. With SCAP, you only need to expand the circuitry
going to the edge of the array.’ To expand systems based on conventional
processor architecture requires extra circuitry on every chip.

Two large Australian defence industry contractors have recently taken
delivery of SCAP systems to evaluate them in military systems, says the
DSTO business office.

A small Adelaide firm, Ebor Computing, is also developing software for
SCAP so that it can be used as part of a system for analysing sonar signals
in the six advanced Collins class submarines being built for the Royal Australian
Navy.

Work on the second stage of the world’s largest hydroelectricity project, at James Bay in northern Quebec, has been held up by the Canadian government’s demands for a new round of environmental reviews. The setback is a victory for the 15,000 Cree and Inuit Indians who live in this vast subarctic region. Their opposition to the scheme is gaining increasing support from conservation groups in Canada as well as the US.

Quebec’s premier, Robert Bourassa, is the driving force behind the James Bay development, which he sees as the path to economic independence for the province. Hydro-Quebec, which is owned by the provincial government, is borrowing US$45 billion to complete the project, which was started in 1975. The first stage of the project, on the La Grande River, began producing power in 1984.

By 2006, Hydro-Quebec plans to have 23 power stations dotted along the three drainage basins which feed into eastern James Bay and southern Hudson Bay. These will produce more than 28,000 megawatts of electricity. Hydro-Quebec hopes to be able to export around 12 per cent of this to the US, bringing in more than $20 billion in revenue over 20 years.

The second stage of the scheme should have begun this spring with the construction of an access road to the Great Whale River. But Canada’s National Energy Board has decided that if power is to be exported then the work must comply with federal guidelines on the environment, not just those laid down by Quebec.

Hydro-Quebec’s appeal against the decision will be heard in June but the matter is unlikely to be resolved until the end of the year.

The delay has averted confrontation with Indians whose fight to stop the project has intensified as the impact of the dam works on the La Grande River has become apparent. Its effects include a dramatic increase in the levels of mercury in the reservoirs. The soil and vegetation over the Precambrian rocks of the Canadian Shield are naturally rich in mercury. When inundated forests decompose, bacteria turn the insoluble mercury into soluble methyl mercury, which accumulates in the flesh of fish. Predatory fish, which build up the highest concentrations of methyl mercury, are a traditional part of the Indian diet.

The Indians complain that their traditional hunting lifestyle is being eroded in other ways because of the new, direct road links to cities in the south. They say that social problems such as prostitution and drug and alcohol abuse are worsening.

Conservationists are also worried about the colossal scale of the dam works. The project requires the construction of 215 dams and dykes, altering the flow of 19 rivers and flooding an area of 26,000 square kilometres. The second stage includes damming the last of the wild rivers flowing into James Bay and southern Hudson Bay.

Jan Beyea, a scientist with the Audubon Society, says that changing the freshwater flows into James Bay could increase the salinity in the bay and threaten food resources for many species that live there, ranging from fish to the many millions of migratory birds that stop off there.

The bay is also home to large numbers of ringed seals and provides a summer retreat for beluga whales. ‘This region is one of the last remaining refuges for wildlife in North America,’ says Beyea.

Hydro-Quebec has indicated that it will go ahead with construction of the second stage of the project regardless of the outcome of its appeal against the energy board’s decision, although it prefers to have the option of exporting its electricity. However, Jacques-Andre Couture, a spokesman for the corporation, says that the delays threaten agreements with power companies in the US.

The corporation has signed contracts with some power companies and is in the final stages of negotiation with others in New York, New England and Vermont. But because the project will not be completed on time, Hydro-Quebec has had to extend the deadline for confirmation of the contracts. This gives the power companies time to reconsider whether they need to import power from Canada at a time when demand has levelled off. It is also giving conservation groups the opportunity to put increasing pressure on the companies to opt out of the contracts.

Power companies are feeling the pinch of recession. They have also found that demand has fallen because of the success of energy conservation programmes brought in to comply with new federal and state regulations.

Just as Hydro-Quebec’s export market is looking less secure, it has found itself at the centre of a major political controversy. This concerns the ‘risk sharing’ agreements signed with 13 multinational companies, mostly aluminium producers, which have set up in Quebec.

Some details of these confidential contracts have been revealed in Quebec’s national assembly in recent weeks. They show that the power corporation had agreed to provide cheap electricity to the companies for up to six years in return for a share of their future profits. At the same time, domestic consumers have been asked to pay much higher prices for their electricity. Outraged consumers suspect Hydro-Quebec is asking them to subsidise these foreign companies.