Navigation path

Policy Image

SEIN - Understanding innovation networks


Turbulent markets and the complex dynamics of science and technology create an uncertain environment for the development of innovations necessary for company competitiveness. Go-it-alone’ strategies are becoming more and more risky in the current climate.

Innovation networks aims to reduce this uncertainty through the co-operation of government, universities and companies. They are a new and increasingly important model for generating new ideas. They bring together, in a formal partnership, the multidisciplinary knowledge that is needed for the creation and development of new products. A computer industry innovation, for example, may involve solid-state physics to mathematics, and language theory to management science.

Economic success depends on this new form of co-operation. Yet little is known about innovation networks. Open questions include: What are the different types of innovation networks and their characteristics compared to classical forms of organisation? What is their structure, coordination mechanisms and the dynamics of how they work? How do they react to changes in their environment? What are the implications of these findings for policy-makers and managers of innovation?

The SEIN project set out to answer these questions. It combined a number of disciplines – science policy studies, evolutionary economics, sociology and computer science.

Its specific objectives were:

  • To establish guidelines for the construction of successful innovation networks;
  • To develop a dynamic simulation model of innovation networks;
  • To provide policy-makers, businesses and the research community with a better appreciation of the likely impacts of their activities in stimulating innovation;
  • To provide new evaluation policy tools.

Work undertaken

SEIN started with an analysis of existing innovation networks in four fields – biotechnology, mobile communication, energy technology and e-commerce.

Using the case study findings, it developed a comparative model to identify common patterns in innovation networks. It then went on to develop a computer simulation model – the General Simulation Model (GenSEIN) – to analyse the possible impact of each actor’s behaviour and the influence of changes in the networks’ environment on performance.

Each actor is represented by an ‘intelligent agent’ whose attributes include autonomy, ability to interact with others, reactivity to signals from the environment, and an ability to engage proactively in goal-directed behaviour. Each actor is given a ‘kene’, a structured collection of technological, political, social and economic capabilities. Kenes change as actors acquire knowledge from other actors and as they refine their knowledge through research and development.

Using the knowledge represented in their kenes, actors produce artefacts, such as a new design or drug, amounting to potential innovations. An ‘innovation oracle’ then identifies the artefacts that will become innovations, i.e. successful new products and processes, and rewards the actors. The oracle maps the artefacts on an ‘innovation landscape’. Successful innovations deform the landscape so that the reward for a second identical artefact is reduced reflecting the real world.

The project also used GenSEIN to develop a new evaluation tool, which concentrates on the comparison of innovation networks and their performance instead of their output.

Key outcomes / conclusions

SEIN made a number of policy relevant conclusions:

  • Communication difficulties exist across the wide range of partners and disciplines. Partners had problems with setting a common research vision and ensuring everyone’s sustained co-operation. The removal of barriers to mutual understanding and a thorough consideration of potential interests and incentives for co-operation are key policy design elements to stimulate innovation networks;
  • A leap in communication, co-operation, coordination and inter-disciplinary skills is required from individuals and even more so from organisations to participate effectively in innovation networks. RTD programmes need to consider this more;
  • Innovation requires creativity, diversity and learning, which implies that the final outcomes cannot be predefined in detail from the start. Policy has to take account of this;
  • While innovation networks need to remain open to new outside sources of ideas, they also need stability to remain on a targeted path. A lead organisation and a clear mission and identity can help keep them on track, with a high degree of trust being a critical pre-condition for co-operation;
  • Actors who in earlier models of innovation only started to matter in the dissemination phase now enter the scene much earlier. SEIN said the role of innovation networks in the market phase is unclear. In fact, many disappear soon after market introduction. For policy, the blurring of the boundaries between innovation and dissemination implies that it is becoming increasingly difficult to distinguish between pre-competitive and competitive research;
  • As it is difficult to anticipate the final knowledge outcome of innovation networks, evaluation should be seen as a support for those conducting the research and not an output control comparing what was originally planned with the final result;
  • As innovation networks depend on many different policy fields, a better coordination of RTD policy with other domains is becoming more important, both in terms of substance and timing of support measures;
  • Finally, new forms of governance are needed, based on the involvement of a broader range of actors and stakeholders in policy design, departing from an experts-only culture.

SEIN shows that the network model may be appropriate for describing innovation processes in many different fields. But it urged caution in adopting a single model. It recommends a differentiated approach to innovation networks in European RTD. Centres of scientific excellence, for example, will require a different model than networks of applied industrial research or networks of knowledge production with a high degree of public involvement.


SEIN set up a website (, published a newsletter and 15 project papers, produced a literature list on innovation research and networks, and organised four workshops and a closing conference.

Publications' list

  • Pyka A. and Küppers, G., ‘Innovation Networks: Theory and Practice’, Edward Elgar, 2002.
  • Ahrweiler, P., Gilbert, N. and Pyka, A., ‘Simulating Innovation Networks’, Research Policy, Special Issue, forthcoming.
  • Gilbert, N., Pyka, A. and Ahrweiler, P., ‘Innovation Networks - A Simulation Approach’, Journal of Artificial Societies and Social Simulation, vol. 4, no. 3, 2000.
  • Kowol, U. and Küppers, G., ‘Innovation Networks: A New Approach to Innovation Dynamics’, Innovation and Regional Development in the Network Society, Technology Policy and Innovation, vol. 7, forthcoming.
  • Pyka, A., ‘Applied Simulation Analysis’, Special Issue of the Journal of Artificial Social Simulation Studies, 2001.
  • Pyka, A., ‘Innovation Networks in Economics – From the incentive-based to the knowledge-based Approaches’, European Journal of Innovation Management, forthcoming.
  • Pyka, A., Windrum, P., ‘The Self-Organisation of Innovation Networks’, Economics of Innovation and New Technology, forthcoming.
  • Weber, K.M., ‘The political control of large socio-technical systems: New concepts and empirical applications from a multi-disciplinary perspective’, New Concepts, Spaces and Policy Tools: Recent developments in Social shaping Research, Edward Elgar, 2001.
Full titleSimulating Self-Organising Innovation Networks (SEIN)
Project AcronymSEIN
Contract number98-1107
Project TypeRP
Keywordsnetworks, innovation, research, S&T knowledge, simulation
Main contractorUniversity of Bielefeld
Institute for Science and Technology Studies
33615 Bielefeld
Phone: +49 521 106 4674
Fax.: +49 521 106 6418
Scientific Coor.Kuppers Gunter
Partners' List
  • Prof. Nigel Gilbert
    University of Surrey
    Centre for Research on Social Simulation in the Social Sciences
    United Kingdom
  • Dr. Paul Windrum
    University of Maastricht
    Maastricht Economic Research Institute on Innovation and Technology
    The Netherlands
  • Prof. Paolo Saviotti
    Institut National de la Recherche Agronomique
    Centre de Recherche d'Avignon, Département d'Economie et Sociologie Rurales (ESR)
    Grenoble, France
  • Dr Peter Fleissner
    Joint Research Centre of the European Commission
    Institute for Prospective Technological Studies
    Sevilla, Spain
Start Date1998-12-01
End Date2001-02-01
EC Contribution€670 000
EC Scientific OfficerM. Carvalho-Dias
Final ReportDownload PDF PDF icon - [149 Kb]