CECS Spotlight: Sea of dreams

A small yellow submarine is making waves.

She lived in a research lab and was subjected to many experiments by her makers. Made from plumbing pipe plastic with some welded fins and heavy wiring inside, Serafina wasn't the prettiest prototype around. But she was well-made and robust and her makers were very careful with her, as she was valuable to them.

One day many months ago, her makers took her to the tank where she often swam. She navigated around the pool as usual. But this day was different. There were people with microphones and many cameras watching.

Serafina was asked to perform for the cameras. She did little dives and changed directions. The next day she was in the newspapers, on television and on websites.

Her makers got many emails and telephone calls about her. People were fascinated by Serafina's skills. Enquiries from around the world came from distributors, potential partners, media and even scuba divers.

Serafina's creators got some new friends to make her cleverer - and prettier - than she had ever been. Now Serafina is set to get a whole school of friends, and travel to some of the most remote places on Earth.

It's not exactly Cinderella, but the path from student project to commercialisation is like a modern-day fairytale for Serafina the autonomous submersible and the team from the Research School of Information Sciences and Engineering who invented it.

Before the media conference in 2004 that exposed the clever underwater robot to the national media, there was little hint of the worldwide interest the tiny submarine would attract. But once word spread about how Serafina could reveal more about the deep sea cheaply and easily, the researchers were inundated with hundreds of enquiries.

"I thought that it could attract a good deal of interest in academic circles," explains PhD candidate Felix Schill. "Swarming and sensor networks became very popular in the last few years, and I also saw great potential in small, non-destructive, cheap submarine robots. These would allow researchers to be more daring with experiments, which can yield greater insights and more surprises without the financial or personal risks.

"I couldn't really predict how Serafina would also make big waves in the public. The first press conference caught us all by surprise, with several big newspapers and the major TV stations present. After that we could track the articles on Google, spreading worldwide."

The team of researchers behind Serafina is now working with a number of partners on components, materials and marketing to deliver a slicker product - Serafina Mark II. The interest in these developments has hardly abated, as the autonomous submersible shapes up to dive in a way never before seen in human machines.

Traditionally, deep sea exploration, mapping and monitoring is an expensive, slow business. Controlling the submersible from the surface is fraught with problems. Meanwhile, if the technology being used to gather information is damaged or destroyed underwater, all information is lost.

Serafina is small, autonomous and flexible. It will be programmed to conduct its operation autonomously without instruction from technicians on a ship. The research team, led by Dr Uwe Zimmer, also wants to keep Serafina relatively inexpensive, so losses are kept to a minimum.

But the most important - and challenging - aspect for the research team will be the development of a 'swarm' communication system, which will allow many Serafinas to navigate harmoniously as a group, much like a school of fish.

"It turns out there are no adequate communication systems underwater," Zimmer says.

"At the moment any set-up requires a lot of power. You can carry that on a military submarine, for example, but in Serafina it presents an interesting challenge because of its size.

"We're working on a communication system that will allow us to organise the school of Serafinas as a single entity, to disperse locally sensed data to all of the submersibles and to grow or shrink the school as necessary," Zimmer says.

Serafina Mark II has been in development since a partnership was formed with Advanced Technology Systems Australia, based near Newcastle. The researchers are now 'burning' Serafina shells in materials testing (the casing has to be tough to cope with the changeable and rough conditions underwater), developing a new 'thruster' (a free-floating propeller powered by a magnet), and working on a successful underwater communication system.

Ideally, a school of Serafinas would swim just like a school of fish, navigating the ocean with perfect synchronisation. The trouble is that it's not really understood how fish swim so perfectly in schools.

"Presumably there is some communication between the fish so they know what move to make when," Zimmer says.

This fishy behaviour has provided a reference point, spurring the ANU team to develop a similarly clever system. The premise of this approach will be to integrate a communication system in each Serafina which would work out communication windows so that no two individual submersibles would 'talk' over one another. By keeping in touch with neighbours in the school, the position and activity of each craft would quickly be known by the entire fleet.

"Our goal is to provide advanced submarine technology to researchers who currently are not able to afford it, and to provide a distributed measurement tool that is easy to use," Schill says.

"The ability to take chemical and physical measurements at different spots at the same time opens up a huge range of applications, most of which we can't even anticipate right now."

Developing a workable undersea communications technology would be a breakthrough and cause as much, if not more, excitement than even Serafina's first splash in front of the cameras. Zimmer and colleagues are hoping to begin testing a swarm of communicating Serafinas next year.

They also hope to make the Serafina cannier by making it capable of critical decision-making. Again, the researchers are looking to the natural world for inspiration.

"We have a long way to go yet on this side of things, but we have been interested in research by some colleagues here at the Research School of Biological Sciences on fiddler crabs," Zimmer says.

"It could be useful to know how the fiddler crabs all at once come up with the decision to run for their holes or stay outside and take the risk in the face of certain dangers."

With teaching and PhD supervision responsibilities as well as the pressures of a commercial partnership, there is plenty to keep Zimmer and his colleagues busy. The team continues to grow, including a number of PhD places funded by commercialisation partners.

The partnerships in place to develop Serafina Mark II impose major deadlines, so the researchers work on relatively short timescales. Zimmer understands why such strict goals are required.

"Is it ideal for anyone to work to deadlines? No, but there is an effect, and that effect is that you work hard and get on with it.

"To be relevant to the physical, practical world, it's imperative we work to deadlines and reach agreed goals. In many ways I enjoy this effect of industrial pressure."

But these deadlines are tempered with an understanding that in any joint venture the technology may not live up to its promise.

"I've been realistic about our limits," Zimmer says. "I've tried not to get too excited because I think you should always expect a flaw somewhere along the line.

"But we're still looking for the flaw in this one. So far it's been promising, and it's comforting to see the research is expanding into many directions - materials science, communications technologies, robotics. Every day I'm more confident that something exciting will come from Serafina."

Schill agrees that the commercialisation process has imposed some tighter deadlines than otherwise might exist for PhD students, but his outlook is positive.

"The 'real-world' scenario is a great source of inspiration and unexpected turns and twists. This can be frustrating for an engineer who wants to deliver a functioning system, but it is paradise for a scientist - you never run out of scientifically interesting, unanswered questions. The commercialisation brings benefits in the long run, even though the work load increases quite a bit."