International Conference on Synthetic Populations: an inspiring event

Our first open conference was focused on the use of synthetic populations in a wide range of fields. We were able to attract world leaders in working with synthetic populations, and participants from CoeGSS mingled with specialists from other centres of excellence. Together with the organisational effort of IMT and the wonderful location and environment, this led to an exchange of experiences and ideas in an unusually creative atmosphere.

In particular, we learned more about:

  • the use of big data on health to develop new categories of co-morbidity that allow for more effective therapies with less side effects
  • the amazing granularity that can be achieved with synthetic populations related to traffic
  • how to harness social media data to identify dynamics for synthetic populations
  • looking at data structures for synthetic populations as social coordinate systems for global systems science
  • how important it will be, especially in Europe, to connect HPC use on policy issues with new forms of citizen participation.

Back-to-back with the two-day open conference a third, CoeGSS internal, day allowed the consortium a deeper look into the technical side of synthetic populations in a training workshop, and then build on the two days of inspiring meeting of minds to sharpen our profile in view of the midterm review.

We identified story lines that we could share with the review team so as to get its highly valuable — and free! — advice on how to build on them in the next months (and years):

  • The health use case is now at a stage where effects of policies on the population dynamics of smokers vs non-smokers can be simulated.
  • CoeGSS can help setting standards for components and interfaces in work with synthetic populations, thereby adressing a critical need of the research community.
  • We are moving towards a seamless process from data to synthetic populations to dynamic structure to simulation to analysis and visualisation, a process that will enable us to create value for a wide array of users.
  • By listening to people from other CoEs it became clear how to further develop our present lean business model canvas in view of phase two of the centres of excellence; this in turn has led to insights about how to develop the portal in view of phase two.
  • To outsiders, HPC may look like a big, static entity, with Europe engaging in a catch-up race familiar from other areas of digitalisation. In reality, the world of HPC is rapidly changing, and Europe can become a leader by preparing for issues that will be vital both for HPC and for society at large in a few years — like the challenges posed by global systems.

Get ready for NetSci 2017 in Indianapolis, Indiana

NetSci 2017 is the flagship conference of the Network Science Society that aims to bring together leading researchers and practitioners working in emerging area of network science and promote interdisciplinarity and collaboration across many areas of network science research.
NetSci 2017 will be held on 19-23 at the JW Marriott Indianapolis in downtown Indianapolis.

High Performance Computing meets Global Systems Science

After the International Conference on Synthetic Populations in Lucca, on these pages we’d like to open a confrontation on the societal impact of High Performance Computing and Global Systems Science.
In the beautiful location of IMT School, Carlo Jaeger, Potsdam University, and Bastian Koller, High Performance Computing Center Stuttgart (HLRS), offered us interesting ideas to think about.

We do not really understand the Global Systems. We have made advances in understanding complex systems at all scales, from molecules all the way up to galaxies. Global Systems have properties of themselves, they are very important for us to understand, it’s urgent and it is scientifically very interesting.

Carlo Jaeger told us, then added: Studying this kind of systems, that’s what we are engaged in.

CoeGSS is an opportunity to bring together people studying Global Systems with people working with extremely powerful computers

“High Performance Computing gives us the potential to go beyond the national limitations of experiments, tests and predictions, we can now simulate things which are too dangerous to do” added Bastian Koller.
If you use HPC, you can make simulations much wider and globally, so you can talk about countries, country borders, you can talk about the distribution of people, the different attitudes, the different behaviours.

The more parameters you can feed into the system the more calculations you can do and with HPC we have the potential to evolve this step by step, to bring in all these parameters and calculate it in a reasonable time frame