Universities joining forces

Professor Reinhard Klette with his students and the Haka1. Photo / Getty Images

A car that can drive itself more safely than a human is among the technologies university academics hope will impress business at an event this week.

The nation's universities have joined forces to showcase their work in 15 expos around the country in the next 12 months, intended to address New Zealand's poor record of co-operation between the scientific and business communities.

Tomorrow night in Auckland information and communications technology academics will display latest research, aimed at selling solutions to business.

A research team led by Professor Reinhard Klette at Auckland University's Centre for Software Innovation will show off a Mercedes A170 that can interpret road signs, speed restrictions and travelling distances via cameras on the windscreen.

Vision-based driver assistance is the hot ticket in the car industry and research groups around the world are racing to help carmakers develop their own technology, Klette says.

His team is partnered with Daimler in Germany.

However the technology has many applications and the team hopes to interest other companies attending tomorrow's event.

It can be used in anything from a boat to a forklift. "Wherever you have a mobile platform and obstacles in the environment to be considered, cameras can be used to understand the features which are not possible to understand using radar or ultrasound or other sensors only."

The event follows a healthcare expo last month.

The next in the series are high-tech manufacturing and energy and infrastructure events scheduled for next month and October.

The programme is backed by Business New Zealand. Chief executive Phil O'Reilly says the analogy he uses to illustrate the historical gulf between academia and commerce is that scientists speak French while business people speak German.

"So what we need is more Belgians, we need more translators. We have to put things in place to enable that."

The Government's Crown Research Institute Taskforce and subsequent research initiatives in the Budget are part of that push, he says, as is Business New Zealand's new website manufacturing.org.nz.

A key feature of the site is the ability for manufacturers to search the R&D, prototype and business services work being carried out in universities and polytechnics.

New Zealanders' number-eight-wire mentality extends to the relationship between science and business, says Will Charles, general manager of technology development at Auckland University's commercialisation arm, UniServices.

New Zealand has "a fantastically transactional approach to everything", he says.

Overseas a collaboration between a research body and a company would be an ongoing relationship, but Kiwis tend to use technology to address a specific problem then move on.

UniServices chief executive Peter Lee formerly worked for the American-based Fortune 100 corporate International Paper.

"I spent US$1 million a year at the Massachusetts Institute of Technology just to access bright ideas from bright people," he says.

Kiwi car gets spatially aware

Kit, the talking car from 1980s television series Knight Rider may still be a futuristic dream but it's one that will become reality, says Auckland University professor Reinhard Klette.

Experts in the German car industry already predict a car that can drive itself on autobahns within 20 years, he says.

In the meantime, Klette and his team are developing a Mercedes A170 called Haka1 (High Awareness Kinematic Vehicle 1) which uses cameras mounted on its windscreen to interpret its environment.

It can recognise stop signs and speed restrictions and alert the driver accordingly.

It also detects lane departures, and calculates the car's speed in relation to the vehicle in front.

Klette's students have identified 210 different New Zealand road signs but it is currently not possible to cover them all.

"However, at university we look into the future."

New Zealand's diverse driving conditions are one reason Daimler is partnering the Kiwi researchers, but the German car company also wants "good intelligent teams" to contribute to the technology.

Klette's students have driven the car for miles, collecting hundreds of thousands of video clips. Different types of cameras and lenses are used to compare results.

One of the next steps is detecting different types of hazards, such as a child running on to the road after a ball. In that scenario there are two moving objects and the technology must distinguish between them.

"So the car can only move in a direction where there is no human being," Klette says.