Research and development (R&D) networks

Research and Development (R&D) networks are a specific instance of collaboration networks, one of our overarching research areas. Here, the nodes of the network represent firms and links represent their R&D collaborations which are usually explicitely announced. Firms can enter or leave the network and links only have a finite life time, during which firms exchange knowledge in order to increase their own knowledge stock.

While most collaborations are bilateral, i.e. only involve two firms that not necessarily have to be connected to the rest of the network, firms can also form consortia, i.e. small clusters in which nodes are fully connected. Most importantly, collaborations come with a cost because firms give up some of their exclusive knowledge and have to invest effort to maintain the collaboration. These costs have to be compensated by benefits from the collaboration, for example joint patents.

This opens several challenging questions about the empirics of R&D networks and their dynamical modeling by means of an agent-based approach. As our publications witness, we address the full range of research questions from strategic network formation to large-scale temporal data analysis. This allows us to address issues such as the impact of cost-benefit relations (for example, severance costs for breaking up or indirect benefits from being part of a network) on the structure of the R&D network. Specifically,  what are the conditions for stability (resulting from individual optimalization) and efficiency (resulting aggregated optimalization)?

Regarding the network dynamics, we explore the role of conditions for the mutual acceptance of a collaboration, such as firm's available information about their counterparties. Different from simplistic models, firms strategically decide about link formation and termination. Eventually, on the aggregated level R&D networks follow a characteristic life cycle dynamics, triggered by innovation booms in specific sectors (e.g. biotech, computer hardware) and their subsequent decline after a few years. Our models are able to predict such structural changes and map them with surprising precision to the conditions of the rise and fall of alliances, on the firm level. The following animated video, which we have produced to show the evolution of a real global R&D network, gives a precise idea of what our models are able to analyse, generate and predict. For more details, please check our publications listed below.

 

Eventually, this research area also contributed to two other areas, i.e. predicting success from dynamical interaction (here, success is measured by the number of patents) and mechanism design to enhance cooperation. In essence, firms in research alliances are not only collaborators, they are also cooperators in the first place. The success of R&D alliances can be only enhanced, but not guaranteed.

Selected Publications

The Rise and Fall of R&D Networks

[2016]
Tomasello, Mario Vincenzo; Napoletano, Mauro; Garas, Antonios; Schweitzer, Frank

ICC - Industrial and Corporate Change pages: 1-30

more»

A model of dynamic rewiring and knowledge exchange in R&D networks

[2016]
Tomasello, Mario Vincenzo; Tessone, Claudio Juan; Schweitzer, Frank

Advances in Complex Systems, volume: 19, number: 1 - 2

more»

Quantifying knowledge exchange in R&D networks: A data-driven model

[2015]
Tomasello, Mario Vincenzo; Tessone, Claudio Juan; Schweitzer, Frank

Journal of Evolutionary Economics (submitted). Available at SSRN and ArXiv.

more»

The effect of R&D collaborations on firms' technological positions

[2015]
Tomasello, Mario Vincenzo; Tessone, Claudio Juan; Schweitzer, Frank

In Proceedings of the 10th International Forum IFKAD 2015

more»

Innovator Networks

[2014]
Tomasello, Mario Vincenzo; Mueller, Moritz; Schweitzer, Frank

Encyclopedia of Social Network Analysis and Mining

more»

The Role of Endogenous and Exogenous Mechanisms in the Formation of R&D Networks

[2014]
Tomasello, Mario Vincenzo; Perra, Nicola; Tessone, Claudio Juan; Karsai, M'arton; Schweitzer, Frank

Scientific Reports, pages: 5679, volume: 4

more»

Newcomers vs. incumbents: How firms select their partners for R&D collaborations

[2017]
Garas, Antonios; Tomasello, Mario Vincenzo; Schweitzer, Frank

arXiv:1403.3298

more»

The efficiency and stability of R&D networks

[2012]
Koenig, Michael D; Battiston, Stefano; Napoletano, Mauro; Schweitzer, Frank

Games and Economic Behavior, pages: 694-713, volume: 75, number: 2

more»

Recombinant knowledge and the evolution of innovation networks

[2011]
Koenig, Michael D; Battiston, Stefano; Napoletano, Mauro; Schweitzer, Frank

Journal of Economic Behavior & Organization, pages: 145–164, volume: 79, number: 3

more»

From assortative to dissortative networks: The role of capacity constraints

[2010]
Konig, Michael D; Tessone, Claudio Juan; Zenou, Yves

Advances in Complex Systems (ACS), pages: 483, volume: 13, number: 4

more»

Modeling evolving innovation networks

[2009]
Koenig, Michael D; Battiston, Stefano; Schweitzer, Frank

Innovation Networks . New Approaches in Modelling and Analyzing

more»

On Algebraic Graph Theory and the Dynamics of Innovation Networks

[2008]
Koenig, Michael D; Battiston, Stefano; Napoletano, Mauro; Schweitzer, Frank

Networks and Heterogeneous Media, pages: 201-219, volume: 3, number: 2

more»