2010 Pearcey Oration

22 September 2010 Melbourne

The 2010 Pearcey Oration was presented by Dr Megan Clarke, CSIRO as part of the World Computer Congress 2010 at the Brisbane Convention and Exhibition Centre, Merivale Street, Southbank Qld 4101, Australia on Wednesday 22nd September 2010

Transcript of Dr Clark's 2010 Pearcey Oration

Dr Megan Clark, CSIRODr Megan Clark, CSIRO

Thank you very much Nick (Dr Nick Tate, Chair of the Qld ACS Organising committee for the 2010 WCC and MC for the evening), and Senator Kate Lundy. It's a wonderful chance to share some thoughts with you, and I absolutely welcome our international visitors and guests to the Congress.

CSIRO was very honoured to have an association with the Pearcey Foundation, and certainly one of our heroes is of course the great Australian ICT Pioneer Dr. Trevor Pearcey, and I will talk a bit more about our history there later.

Many of you in this room, have the difficult job of looking into the future as I do, and when I want to look into the future and really understand, I go and talk to our astronomers. If I want to really get a sense of the future of communication, the future of data handling - even the future of how we do science, I go talk to our astronomers, because space in a sense of communication is the ultimate wireless. One of the absolute gems, of course, is now Voyager 1 - is really the ultimate wireless system. It's now the most distant spacecraft from earth, and we are still in weekly radio contact with Voyager 1. It was launched in 1977. It's 17 billion km from the Sun. It is twice as far out as Pluto and we are still in weekly communication with Voyager 1. NASA tracks Voyager 1 and 2 through the Deep Space Network, and of course CSIRO managing the Tidbinbilla Station, which is the station, of course, that has the Southern Hemisphere part of those three stations.

Have you any idea how much power little Voyager 1 is using? It operates on less than 20 Watts of power. The light in your refrigerator has more than 20 Watts of power. There was a few months... you can imagine how attached our dedicated team at Tidbinbilla are to Voyager 1 and Voyager 2. And a few months ago, they had a bit of a scare because they lost contact, and you really don't want to lose contact with this amazing piece of human engineering, but they did a few manoeuvres and gave it some instructions and got it back online, and they are still talking to both of those spacecraft.

I think it really underlines for me, as I talk to our communications team and they put up lots of fancy equations, but at the heart of every one of those equations are two very simple things that even I can understand, and that is the signal to noise ratio. And our astronomers who manage deep space communication, they are defining the future of our wireless communication, because it's unbelievable what they're listening to, and how they're managing to hear those faint whisperings that we have from space.

Who would have known that our investment back in the 1970s to allow our astronomers to listen to some of the whispers of black holes, and they really wanted to hear the radio signals coming from black holes - who would have believed that that was actually the foundation for what is now the ubiquitous wireless LAN technology currently used in a billion devices around the world, projected very shortly to be in 4.5 billion devices, all because of something that happened in a laboratory in Australia.

Four of the five scientists whose names are on the patent, Graham Daniels, John O'Sullivan, Terry Percival, Diet Ostry, John Deane came from radio-astronomy.

But our vision is expanding. Almost 20 years ago, scientists around the world had a vision of a telescope so large and so powerful that it could answer some of the most fundamental questions in astronomy and in physics, and that vision was the beginning of the Square Kilometre Array Project. The Square Kilometre Array Project will be the world's most impressive telescope. It will be capable of looking back right to the dawn of time, and my astronomers assure me it will be able to see every single super-galaxy that exists, and I believe them.

Of course we have two pilot projects for the SKA, and the site for one of those is here in Australia, and that site in Western Australia is as close to radio quiet as you can possibly get. It's in Murchison and the Shire of Murchison, and for our visitors it's an area of 50,000 square kilometres, and it only has 110 people, and that's a population density of just over 2,000ths of a person per square kilometre. I think they even did it in parts per million.

But what the team has already done we've got the first antennae already up, and as Chief Executive it was on time and on budget which my team knows means a lot to me. It doesn't mean much to you but it means a lot to me. But the exciting development here was they linked the very first telescope up... the first antennae up with our radio antennae's at Parkes. For the Australians in the room, "The Dish" at Narrabri, at Tidbinbilla, at Hobart, at Ceduna, at Walkworth in New Zealand. And they created for the first time one giant telescope, 5,500 km in baseline, and that is now available to all scientists around the world. It has ten times the resolution of the Hubble Telescope... in the radio spectrum.

One of the first projects that the team did was to look into the heart of a galaxy called Centaurus A. If you and I could see through the eyes of our SKA pilot project, and we went out tonight and looked at the sky, just under 1/5th of the sky we would see wonderful structure of radio waves emanating from Centaurus A - beautiful arching structures coming out of the poles, and it would take up nearly 1/5th of the sky. It's there, and ASKAP, of course, can see it.

Right at the centre of this galaxy and lurking in this galaxy is a black hole that emits jets of radio-emitting particles close to the speed of light, so the teams, with this new 5,500 km telescope observed this galaxy for ten hours. They took enough data to fill a stack of DVDs in their cases as high as a nine story building in ten hours, and this data was used to make the first image of that tiny region at the centre of this black hole right in the middle of this super galaxy. It's 14 million light years away from us, and it was like taking a photograph of the head of a pin from 20 km. You can sort of see why I talk to astronomers when I want to understand the future of communication.

But they are not just pushing the boundaries of communication they are also pushing the boundaries of data handling. The demands are extraordinary from this telescope. In the first six hours of operation, it will generate most data than the entire history of radio-astronomy and in a week, as much of the entire internet has today. And of course, when you look at the capital budgets, that's an awfully big super-computer and that will be the Pearcey Supercomputer in Western Australia.

Not only are we trying to keep it at the forefront of data handling, but also we will use renewable energy, both for cooling the super computer, and renewable energy at the site, which will be a world first.

But one of the other things that we are seeing is that this is the way that science will be done, and that is the collaboration. Going out to the communities around the world and saying, "We will have this new facility available. Would you like time? What are the projects that you would like to do?" We were five times over-subscribed. There's 12 years of time booked already on this facility, from 160 institutions, from 23 countries - that's what the future of scientific collaboration looks like and that's what they're already doing today.

So, for those of you who are contemplating the end of the supercomputer, I can assure you that our human quest for discovery will simply drive more and more information and not less. You can sort of see why I love my job.

I thought there was a couple of other things that I should share with you that may be useful. For those of you that are looking for where are the hotspots of, perhaps, new revenue, of new opportunity, of new markets, and one of... CSIRO has the largest investment by external partners in Australia in R&D. And we do see across nearly every one of the sectors, where our partners are investing their money in future developments and future services, and there are just two areas that I wanted to really pull out that we are absolutely seeing. Increase, very quickly, in terms of investment and new product potential, new industry potential, new potential for jobs.

The first is of course, the rise of innovation in a service industry. We are certainly seeing a wave of innovation around personalising services. An example is we have a new project now with Centrelink. For the Australians, they know what Centrelink is. For our visitors, it is... it sees a third of all Australians and provides services to a third of all Australians in terms of social services and support. And of course, that service needs to be delivered with absolutely the best mathematics, statistics and network science to be able to provide for the individual needs of the users. And it's certainly going to stop... what Centrelink wants to do is try and identify all of that information and try and find the trigger point and the traps so that people don't fall into poverty.

So taking exactly the same technology that we are using in space, and now using it to make sure we help people get a job, and stay in that job, and not move into poverty.

One of the other areas that is emerging for us is this hotspot of the network of things, connected with each other - more and more devices communicating with each other. There's obviously challenges around that. How do you get heard in the crowd of a network - something I think my kids are becoming expert with on Facebook. But how do you ensure reliable communications? How do you make sure you have unique addresses? These are the things that I'm sure the people in the room can answer very well.

But we're learning that we don't understand a lot about these networks - that they way they connect are neither purely regular, and nor are they purely random. We also know that our mathematics is not good enough to describe them. And what we are learning is that these complex networks - scale free networks; small-world networks; major discoveries and these breakthroughs are influencing our science and discovery in mathematics, in physics, in biology, in computer science, in sociology, in understanding complex systems and trade-offs in very complex worlds. So, this is an area that we are seeing more and more investment.

One of the most sobering conversations I had in understanding networks, and the country that this request came from will remain confidential, but I had a request from the head of Police of a country who said, "We need to keep ahead of the sophisticated networks of some of our drug dealers". And I'm thinking, "I am not sure this is a project we really should be doing". But he said, "We can see them, because of their mobile phones", and we know that apparently if you are going to kill someone in the drug world, you phone them first to check that they're home, and then you drive across, and they could see them driving across, and then they could see them. So they could witness what was happening in the network and they could see all the players and what was happening, but they realised they needed to understand this. So it's in many, many sectors that this deep understanding. So anyone out there in the audience who is a network scientist, you are in for a roller-coaster.

I also thought many of you, like me, need to manage the complexity of how do you make sure your organisations have, and your people have access to the best tools, the most advanced tools, that they can stay on the cutting edge, but you also need to manage efficiency and cost - the reality of that. So I thought I would share what we're doing as an organisation to balance that, and what we're doing, particularly with our Cloud Computing. So whilst I mentioned the need for supercomputing, and we will continue to do that. We have over 6,500 researchers in CSIRO - 20-30 of those researchers consume 90% of our supercomputing resources, and yet we need to make those sort of resources available to all of our researchers in the future.

One of the things that we are doing, of course, is looking at Cloud Computing, and looking at how we trial and use that. Already we have 1,000 desktops linked in grid-computing, and already we're making that available to some of our researchers. That provides us 2-300 times speed increase and we are trialling a lot more in this area. So, I just wanted to share that because I know many of you will be making the decisions and are probably interested.

But I mentioned the future, and mentioned some of the hotspots of science and technology that we are also exploring, but our mission at CSIRO is to make sure our science is used. Make sure it is used in the economy; it's used to build new industries to support our existing industry; it's used to improve the health and wellbeing of all Australians and it's used to create a sustainable future for our environment. More and more we are also seeing that a world where Australian science, or our science can help the world develop a sustainable future for all of mankind and all of human kind.

And so our mission - we don't get a tick in the box until a science is out there being used in the community, and to deliver that kind of profound impact for the nation we absolutely need our partners to collaborate, and to bring those technologies alive and bring them out into the community.

One example right here in Queensland, and I know Phil Gurney is in the audience, and this is some of the research of his team. We're working with the Queensland Government and Queensland Health for the Australian eResearch Centre here, and that aims to improve the quality and safety of healthcare for individuals and their communities.

One of the most interesting projects that this team is doing is how on earth do you computerise the health records that come from all different areas, all different bodies of information. It sounds so simple just to digitise the records, and of course it's incredibly complex and the team has done a world first in being able to recognise and categorise clinical terminology, no matter where the record is and no matter what form it was in, and that is extraordinary. Those kind of little steps are so important if we are going to reform the right records.

Another area that is making a huge impact is that in our hospital service we balance elective surgery, with, of course, our emergency services. And elective surgery, in most areas, has higher margin, but you absolutely need to have your resources available. Who would have thought that by using the latest in mathematics, and by using the smarts and creativity of our researchers that they could get very good predictive tools about how to predict what happens in an emergency centre, and in fact that's exactly what they're doing. So, they're now forecasting daily, four-hourly, hourly intervals predicting the gender, the admission type, what sort of injury it is, whether it's orthopaedic, cardiac - this actually can be predicted and in an emergency department can be predicted in a way we never imagined possible.

This is now in over 27 public hospitals in Queensland. Queensland is leading the way in this, and the dollar savings and efficiencies are significant. Even for a single hospital $15-20,000 saving per day, but most importantly, people are getting better healthcare.

Other examples that we're doing, of course is tele-health - diagnosing eye diseases in our indigenous populations, and certainly using much better diagnostic tools to connect people in Western Australia and some of our remote areas. All of that, of course, requires broadband, and it means that we need to have broadband through 100% of Australia. One of the projects that CSIRO is working... [Applause] We are committed to that.

As the national research agency, we actually believe that we do have a huge commitment to provide that last 10% and our researchers are working unbelievably hard to do so. Our aim is really to make sure that that wireless broadband is delivered so that we can have new health, government and educational services to all Australians.

We are trialling this technology in Smithton in Tasmania in partnership with NBN Tasmania. It's a very important first step to get this technology that we have developed in the lab out into the community, and demonstrate its potential and cost effectiveness, and I know already there has been further discussion at this conference in terms of how we better use that.

The solution is an elegantly simple one. It reuses our analogue television spectrum. We've developed completely new antennas, and they're mounted on the existing towers, so it's a very cheap and effective way of doing that. We are targeting a very aggressive timetable, and that is making sure we can really deliver on this by 2012. And I commend you to the work that is being done. Ian Opperman who is here is leading that effort, and his team is working very hard.

I wanted to close with just a little bit of history and talk a little bit about Trevor Pearcey. It was the 1940's - CSIRO scientist Trevor Pearcey, Maston Beard, Geoff Hill - they built a digital computer called CSIR Mark I, and this computer could store several kilobytes of data, and it was the fourth computer in the world. Steady improvements, they tell me, on Mark I through to the early 50's happened in Sydney, and I think it culminated... apparently the peak of this computer was when it buzzed the world's first digital music - Colonel Bogey of all things, to an international audience at the first computer science conference in 1951. I think that was the peak of its career, actually.

And I think Cuthbert Hurd at IBM was scratching his head at that time going, is there really a market, and I think he predicted that they might find as many as 30 customers for these new computers. And CSIRO at the time also was looking at where it should invest its money, and made the decision, at that time, to get into radio astronomy and to move into the wireless side of things. So Mark I apparently had a very humble, I suppose, ending, in a way. It was dismantled and driven by car down the Hume Highway to the University of Melbourne on the back of a truck and it's now on display in the museum, and it hasn't operated since 1964. But that was a very pivotal decision, and it embarked Australia, and CSIRO, of course, on the endeavour of radio astronomy that we are now progressing very, very aggressively on behalf of the world.

So last year, as you know, CSIRO was lucky enough to license Wireless LAN to half of the laptop market. For those of you using the other half of the market, you should shift. We have now set up, with the proceeds of that a Science Industry Endowment Fund. That money is the money we're already using to help fund our rural wireless broadband research to bring wireless broadband to all Australians. We're also using it to provide a Professorial Chair in Wireless Communication at Macquarie and established scholarships and fellowships for young Australians in ICT, Mathematics and Engineering, and a very important reinvestment back into this area.

We are very proud to be associated with the Pearcey Foundation, and we are absolutely delighted to be sponsoring for the first time next year, the Young Achievers Award. And I really look forward to announcing that, because I think as we've talked about the future being space, but really the future is our young people, and making sure that we're making the sort of investment in them, so that they can absolutely maximise discovery in ICT, and maximise the impact that that technology will have on humanity.

Thank you very much.