Crunching the numbers for the common good

Melbourne is Australia’s fastest growing city. Experts estimate that by 2050 its population will reach 8 million. How will that number be managed, housed, fed, watered, supported and employed?

The Australian Urban Research Infrastructure Network (AURIN), a spatial intelligence network housed at the University of Melbourne, is working on that awesome issue. It is measuring our nation and guiding its future.

Dr Serryn Eagleson leads the team that collects huge streams of data from myriad sources, analyses it and delivers the results to researchers and policymakers charged with advising governments, organisations and companies on what lies ahead, and why.

We put five questions to her about how big data can be used to inform some big decisions.

1. The data is obviously valuable and unique. Where does it come from, and who wants it?

We have about 80 different data providers – the Australian Bureau of Statistics provide the Census, of course, plus scientific organisations, industrial peak bodies, companies and so on.

Increasingly, too, data is supplied through citizen science. People have mobile devices, handheld detectors that measure such effects as atmospheric pollution levels, and they feed that information to us. For instance, data provided in this way helped reveal that some cases of lung cancer in non-smokers had been caused by pollution.

We have 1850 data sets that have been curated and are accessible through our website. We are funded by the Federal Government through the National Collaborative Research Infrastructure Scheme (NCRIS) and housed by the University.

We don’t charge for our information. Anyone can see it, but a custodian controls access and some data are secure. There are sub-sets for specific users, such as researchers who apply to the custodian for approval and permission.

2. Do you know who is accessing your information?

We know that 7500 people have accessed the system. We know that 84 per cent of them are academics and 10 per cent are governments, and we know the kinds of information they are getting and who they are in the sense of their department, organisation or research establishment, but we don’t know the individual.

Many councils use our data. For example, we contributed to the Resilient Melbourne project that looked at how pockets of isolation developed and poor health and other social, physical and economic challenges.

3. How do you present the information so it can be easily understood?

It is based on statistics, mathematics and geometrics, but I also look on it as an art form. We present data as maps of the location under study, working on computers to visualise it in various ways, and so it can be examined according to various criteria.

For example, a project to identify and help isolated, lonely, elderly people used our data to map where such people live. This, overlaid with other maps, such as one indicating where government and other services are available like planned activity groups, volunteer visitors and so on - it showed what the future might look like. It was also about how older people might be kept working, stimulated, active and contributing.

The project may be about siting a new health facility, so it’s about understanding the demographics, transport linkages, spatial information and council plans for that site. We did a land use study of the railway lines going through Melbourne to understand their capacity for growth and made recommendations based on that.

A lot of our work is about where people live, where they get their first job, where their local connections and network are.

The critical thing AURIN has done is to make accessible data that is meaningful.

4. How do you keep such a large volume of information up to date?

We constantly update the data we have. As well, we decide on the type of data we want and then go looking for someone to provide it. We point out that it is for research and the public good. When we get the data, it has to be structured so that it can be read by the system.

Then we monitor its use; how many people log in to use it. Is it valuable? If it is, we update it. If it is not being used, we archive it.

We are dealing with quite rapid changes in population so that requires constant updating. We can have live information but we also have archival information so we can see how something has changed, over time and over areas. Changes such as the gentrification of an area – meaning an increase in disposable income and changes in the interests and activities of the residents – are important to identify.

We can also track the movements of migrant populations as their fortunes change. We can track the movement of pockets of people and also see if some ethnic groups are clustering together.

That’s important for purposes such as matching available services with the needs of specific populations. We work with the research community; we provide the data and they are the experts to do the analysis. We have built a whole range of case studies – housing affordability, use of hospital emergency services, and so on. The range of studies is very wide; for instance, we did a case study on live music scenes and what stimulates them.

We also look at how the population of cities will grow in the future, making a computer animation on a number of assumptions and putting it against a timeline to see how things might change.

We have done a study of what Melbourne could look like in 2051; where people might be living and how it might mesh with fringe developments that have already been planned and how we will deal with autonomous vehicles that require millimetre accuracy in control.

That could have immense consequences for the planning of infrastructure.

5. When did your interest in this kind of work begin?

At AURIN, I am Manager of Data and Applications and also the Assistant Director of Research at the CRC for Spatial Information. I grew up on a cattle station in East Gippsland but I was always interested in mathematics, statistics and data.

My father was a builder. He spent a lot of time surveying land and I spent a lot of time out there with him. Geography, geometrics, spatial information are where my interests lay, and still do. So I moved to Melbourne for a university course in those areas. My undergraduate degree was a Bachelor of Geomatics/Art and my PhD was in geographical information systems, both from the University of Melbourne.

I was interested in the analysis of that kind of information but also in the artistic side. A lot of what we do is representation and visualisation, working out the right way to present such information. Maps and cartography are art forms in themselves.

My training was all about spatial information and how to bring that alive for improved population specifics, and understanding where people live and how that impacts their lives and how it can be changed for the better.

So we think of green spaces, what services are needed, whether people can access their health services on foot, and how, as a society, we create liveable spaces.

A version of this article also appears in the University of Melbourne’s 3010 Magazine.

Banner image: iStock