No time to question the need for speed

The Government paid $43 million for New Zealand to join the club of countries operating high-speed research networks when it decided to meet the capital costs of the Kiwi Advanced Research and Education Network, or Karen, in 2006.

The network was turned on 20 months ago with the aim of facilitating research between universities and other research organisations in New Zealand and around the world. Next month, Cabinet is due to review its business model and see whether taxpayers are getting their money's worth.

For some, the investment was a matter of national pride. That was much in evidence earlier this month when Research and Education Advanced Network New Zealand – set up to operate the network – for the first time played host to the biannual Asia Pacific Advanced Networking conference, drawing 300 cyber-scientists to Queenstown.

Karen can transfer data between New Zealand institutions at speeds of 10 gigabits per second using fibre supplied by TelstraClear. A 620 megabit per second link to the US and a 155Mbps connection to Australia are both supplied on the Southern Cross cable.

Reannz chief executive Donald Clark says the network is currently running at an average of between one and four per cent of capacity, but he is pointedly unembarrassed by those figures.

"Great, isn't it?," he says, with a flicker in his eyes that suggests he might break into a haka in front of the Japanese computer scientists jugging plates of noodles during a break in the conference proceedings, if challenged on the matter.

The whole point of a research network is to have plenty of unmetered bandwidth available on demand, and if the network ever averaged 20 per cent capacity it would be time to upgrade, says Mr Clark.

Waikato University head of computer sciences Tony McGregor agrees. Though he admits usage is "a little lower than what you'd hope for in the medium term", he says it is not unusual and an equivalent network in the United States, Abilene, only recorded a few per cent capacity at first.

He says a "massive" international trend towards e-research, where individual researchers around the world collaborate and work as a team using Internet networks, means Karen is a vital resource. "For us to continue in active research, it's essential for us to join these teams. I would even go so far to say that any university that didn't continue to use Karen was putting its research future at serious risk. just because of the international trend towards e-research."

Otago University IT services director Mike Harte also says it is positive there is "plenty of space" on the network for people to use. An Otago University physics lecturer uses Karen to hook up to global lightning sites around the world for real- time observations of lightning, and the network is also used to provide a video link between deaf students attending lectures and a sign language interpreter in Auckland.

"We're also using Karen for access grids – which is video conferencing on steroids. It's a room with high quality cameras and multiple data projectors to provide a collaborative environment. "Multiple sites are linked together in a virtual room and we're able to talk with people in New Zealand and all around the world."

Reannz is nevertheless touting for new customers and utilisation of the network is likely to rise as schools, libraries and even some businesses are allowed to hook up.

Auckland technology firm Endace will become one of the first commercial companies to join Karen when it opens a new research and development centre in Auckland next month. Chief executive Mike Riley, who has become a director of Reannz, is in no doubt as to its value.

Spun off from Waikato University to market software and equipment that researchers had built to analyse the performance of high-speed networks, Endace is a "poster child example of what happens when the education sector and the business sector get together", Mr Riley says.

One of the country's most successful high-tech firms, Endace is now valued on Britain's AIM stock market at £64.2 million, employing 65 people in New Zealand and about another 20 outside the country.

It expects to take on another 30 staff in Auckland by the end of the year – encouraged in part by new tax breaks on research and development expenditure that could be worth as much as US$1 million to the company this financial year.

THE FACT Karen is a closed research network means it is intrinsically secure, says Mr Riley, and the firewalls and caches that slow down corporate networks can take a back seat to raw performance. "The only thing that is going to slow this down is if the users can't articulate why they need this."

Some schools will also begin trialling the network this month. St Mary's College in Wellington will be one of the first to get a connection and it is expected most schools on the Wellington Loop – a local open access fibre network – will be hooked up.

Karen will allow French and Italian television channels to be streamed to language students, and the remote delivery of courses, says Mr Clark.

The APAN conference was not for the squeamish.

The highlight of the event for many of the assembled technologists was a series of surgical operations that were streamed across the region and shown in full gory detail on a 34-million pixel ultra-high definition display, constructed from 15 high-definition TVs.

Christchurch doctor and telemedicine pioneer Stuart Gowland says plentiful, low-cost bandwidth provided by research networks such as Karen allows surgeons to view and provide advice on operations as they are performed.

He says operations in future could be overseen by surgeons from national or global centres of expertise. "We are on the cusp of a major increase in collaboration. Medicine has always worked like that. It is just that we are now on the edge of collaboration being much easier – that hospital corridor can now be a world corridor."

Reannz would seem to be walking something of a tightrope expanding the uses to which Karen is put, while maintaining the network's research ethic. And applications such as ultra-high definition telemedicine, that might be experimental today, could be business-as-usual in years to come.

TelstraClear and other telcos which tendered to supply bandwidth on their fibre networks for Karen were prepared to offer unmetered capacity – technically speaking a wavelength of light on a strand of fibre – at a low cost for academic research.

But Mr Clark concedes it is possible they might have a different view on appropriate pricing if organisations could use Karen for general Internet access and file sharing, as this would risk cannibalising telcos' revenues.

He speculates a hybrid model might be developed for a "Karen 2.0" that allowed both activities.

For now, Reannz' relationships with telcos and the Government appear sound, and its challenges technical rather than political.

Treasury papers indicate that the Government plans to allocate the $15 million it earmarked in the Budget for the support of a new trans-Tasman communications cable to compete with Southern Cross to Reannz, allowing it to become a foundation customer on a new cable that Reannz would select through a tender. This would suggest Reannz has little to fear from the impending Cabinet review.

State-owned enterprise Kordia is considering building a cable in partnership with Australian firm Pipe Networks and is odds-on favourite to be the ultimate recipient of the government grant. But TelstraClear and even Telecom could be rival contenders. Mr Clark won't be drawn on whether he expects Reannz will find itself choosing between multiple bids.

Telecom's Chorus was chuffed at being named a "silver sponsor" for the APAN conference in return for linking the Queenstown conference venue up to Karen, and Mr Clark says the biggest challenge facing Reannz is conquering the "last mile" – making sure it can deliver the full benefit of its bandwidth to researchers' desktops, which was the focus of many of the highly technical presentations at the Queenstown conference.

"What I get concerned about is when people try to limit infrastructure investment because they are worried whether anything good might happen," he says.

"We seem to spend $1 billion on roads each year without worrying about it. Why? Because we have had them for decades and know how to get value out of them."

- © Fairfax NZ News