Global Systems Science for Urban Development

In mid-February 2013 I took part in a two-day workshop at the European Commission in Brussels on Global Systems Science for Urban Development.  By Global Systems Science (GSS) we mean the application of systems thinking to a broad range of global challenges in climate, finance, energy, and other domains.  See the introduction to this meeting for a better overview.  Generally this implies conducting research to understand the many interacting systems within each domain, to collect data representing these systems, and to develop analytic methods and “models” of these systems.  The goals of this work extend from practical tools for decision support in the field to evidence for policy making.

The meeting included urban planners, urban economists, EC policy makers, Big Data analysts, and Lisa Amini of IBM’s Dublin research lab and myself as engineers; see list of participants.

This meeting was one of a series growing from an inaugural meeting in Brussels in November 2012.  The series is linking up with interested parties across Europe, the United States, Canada, and China.  In Europe it is mainly driven by Ralph Dum, who is a scientific officer with the European Commission in the DG Connect.  His goal in exploring GSS for Urban Development – as well as a several other application areas of GSS – is to develop a research agenda in these areas for the EU’s next research programme.  However the operation of the GSS initiative is managed by the Global Climate Forum.

The programme of the GSS meeting is posted and I will not attempt to go through these, but will give a few brief comments:

  • The word “model” was a source of confusion as it can mean several different things.  I also suggested that the group needs to develop a common terminology to enable brevity and specificity in discussion and publications.  There was perhaps an overconfidence in the ability of models to reveal insights.  In subsequent email discussions I have pointed Ralph and others a) to the dictum “All models are wrong.  Some models are useful” and b) to John Sterman’s article “A Skeptic’s Guide to Computer Models“.
  • Interests were concentrated around academic research issues, rather than solving real-world problems.  This lead to some tension between the volumes of data and complexity of simulations required.  Researchers felt that they needed even more data (Big Data) in order to completely represent complex urban systems.  I made the point that in the real-world, we are usually concerned with enabling a manager or a citizen to make a choice among a small number of possibilities and so completeness is less of an issue. 
  • Lisa Amini made an excellent observation that today IBM generally build simulation or prediction tools from basic principles for each new client using PhD analysts supplemented with skills in urban systems.  This makes it not a scale-able business.  Ralph pointed to CAD systems and their templates or patterns that enable lower skilled professionals to deal with architecture, engineering, and planning models.  I reminded the group of the famous book “A Pattern Language” by Christopher Alexander, et al. which had defined patterns for architecture and urban design, but which had been ignored by these professions, although it was highly influential in the development of object-oriented programming.  I suggested that field work to determine the large number of common patterns across cities would be an excellent research area.  There was discussion about the need to include patterns of human behaviour in this research, including understanding how people make choices and how these choices may be influenced by giving people additional views about the impacts of their decisions.
  • It appears that (in Europe) the idea of a Smart City is widely believed to have something to do with embedding lots of sensors and collecting data, but with little idea of what purpose this solves.  GSS would like to address that deficit.  I used a phrase – “making the invisible visible” – in connection with the role of information in Smart Cities, which was much liked by the group.
  • Towards the end we had some discussion of how the application of GSS to urban systems can be established as an academic discipline.  I gave the examples of Computer Science in the 1960s and more recently Services Science as new cross-disciplinary and inter-disciplinary fields that have gained recognition.

There will be a summary of the meeting that I think David Tabara will write up.  The output from the workshop will contribute to an Orientation white paper that Ralph Dum is developing as input into the Horizon 2020 programme.  I was asked to write this section on GSS for Urban Development.

The Brussels workshop is followed by a related workshop at Arizona State University in Tempe where I am at present.  I will write this up in a few days.