Environment and Ecology

What is complexity science?

Complexity science is a broad and multi-disciplinary subject. In a wide range of systems that are the subject of study in biology, in the social sciences and in industrial applications, computational modelling is undertaken to study the behaviour of these sytems; Mathematical developments and modelling approaches from physics can be used to better understand these systems; And expertise in domains from software engineering to systems biology can be used both to inspire new approaches and apply new results.

 The concerns that complexity science addresses has developed from investigations from a varied intellectual ancestry. Some of it has developed from work in cybernetics in the 1940s, to work on general systems theory in the 50s, chaos and catastrophe theory in dynamical systems in the 1960s and 70s, to work on complex systems spearheaded in the 80s by groups like the Santa Fe Institute. Some of this work focussed on abstract mathematical systems and simple physical systems, e.g. sand piles, but more recently, interest has increased in complex adaptive systems, such as social systems, biological systems, and technological systems where the parts actively change the way they interact. The increased use of computer simulation and interest in biological questions created research in artificial life and the simulation of adaptive behaviour in the 1990s.

  Now, in the 00s, complexity science takes in parts of all this rich background of work. An important part of the current emphasis of complexity science is its application to practical technological systems; industry needs to know how to design, manage, build and control systems as they increase in size and connectivity. They want to be able to build systems that are scalable, robust, and adaptive by using properties such as self-organisation, self-adaptation, and self-repair that biological systems utilise. Thus the contemporary applications of complexity science are complemented by a rich background of theoretic work, and continue to address deep scientific questions about nature.Complexity science is a subject of study that is in the perfect position of bringing together deep scientific questions with application-driven goals across many interesting domains.

 Complexity science touches on almost all aspects of modern technology and science creating lots of exciting new opportunities in training, careers and research. One thing to be careful of: ‘complexity’ doesn’t just mean ‘complicated’. Complexity is not just determined by the number of parts a system has, for example. Engineered systems, such as a microprocessor or the space shuttle, might have a very large number of parts with very intricate design. But complexity science is interested in dynamical properties like self-organisation, adaptation, and emergence. Often engineered systems are designed to minimise these tricky dynamical properties – they can make the system difficult to design, predict and control. However, if desirable emergent behaviours can be harnessed and exploited they can help us to move beyond the limits of traditionally-engineered systems that are merely complicated.

Why is complexity science important now?

The rapid increase in interest in complexity science is being driven predominantly by new challenges and demands in technology. Various industries are becoming increasingly aware that traditional approaches to design and engineering are failing to keep up with the increasing scale of today’s systems.

"The management and design problems facing modern ICT [information and communication technologies] practitioners are critically concerned with ensuring reliability, usability, robustness, efficiency, effectiveness, security, and evolvability in the interconnected ICT systems upon which societies and economies increasingly rely. As our world becomes an ever more interconnected place, so-called “systems” ideas and perspectives become increasingly important. A central issue is the emergent behaviour of complex systems.'' - UK government's Foresight report on ICT and complexity. (Read a non-technical summary.)

Systems from telecommunications, to information storage and retrieval, to economic trading, are rapidly increasing in scale. Also, processes and transactions that used to be implemented manually are becoming automatic, and many systems are being connected together. These increases in scale and connectivity make managing the complex dynamics of such systems difficult. Traditional approaches to engineering try to remove complex dynamical behaviours and emergent phenomena that are difficult to control and manage - but they often happen anyway, producing errors and undesired behaviour that can bring the system crashing down. New approaches to design, engineer, manage and control complex systems are urgently needed. This produces a demand from industry to both drive new research at universities and seek new graduates trained to understand and deal with complexity. This creates an urgent need for a change in training:

"Open challenges that must be faced by the complex systems community include overcoming institutional and cultural obstacles to interdisciplinary and industrial involvement in complexity research. In the UK in particular, most computer science undergraduate degree programmes currently have a manifest lack of formal training in complexity ideas and techniques: especially simulation modelling methods, experiment design, and statistical data analysis.'' - ibid.

This need for new training in complexity science has now resulted in the development of new courses in complexity and related subjects, (see Courses).

Another driver in the recent upsurge of interest in complexity science is the availability of computing resources with sufficient power to model large scale complex systems and investigate new ways of approaching their design. Without this, using only formal mathematical methods and traditional design approaches, the design of large scale systems was forced to avoid complex dynamical behaviours because they were too difficult to predict. Computational modelling allows new approaches that were not previously testable.

The third main driver is systems biology (or "the new biology", as it is sometimes known). With the advent of high-throughput devices (e.g. genome sequencing and microarray data) the biological sciences are for the first time able to gather information about whole systems and begin to ask questions not just about one detail at a time, but about the complex interaction of the components together. There has been an enormous change in the way cellular and molecular biology, in particular, is being researched, and large amounts of funding from the biological research council is being directed to these new approaches (see Bioscience for society: a ten-year vision, BBSRC (2003) and UK Biosciences: The next ten years, BBSRC (2004)).

Complexity Science interfaces with systems biology in two different ways:

1. As systems biologists gain new information about how biological systems work, and how they cope with and exploit emergent systems level phenomena, we gain new insight and inspiration for tackling complexity in engineered and technological systems.
2. As we tackle complexity in technological systems and gain understanding of new approaches to modelling and controlling complexity in engineered systems, we can provide new insight that can help biologists understand complexity in natural systems.

Where is complexity science headed?

One of the most exciting aspects of complexity science is its interdisciplinary nature, and the interface with the life sciences is paramount here. Biology (from ecology, to organismic biology, to neurology, to cellular biology, and molecular biology) is filled with marvellous examples of complex adaptive systems that not only cope with emergent dynamical behaviours but have adapted to control and exploit them in every way imaginable. A lot of research in complexity science is looking for ways to model, understand and extract the useful properties of biological systems. This is both with a view to better understanding of the biological systems systems biology and for inspiration for new approaches to solving technological and engineering challenges.

In the figure below, the left side lists some biological complex systems, and the right side list some example systems from ICT (information and communication technology) that need new approaches to handling complexity. The topics in the centre are examples of subjects that help connect the biological inspiration on the left with the challenges on the right.

What opportunities are there in complexity science?

http://www.complexity.ecs.soton.ac.uk/index.php