Using computer models to understand systems

Our researchers use computer models to improve their understanding of complex systems. Sustainability in the meat production chain is, for example, the subject of a new model produced by LEI and Livestock Research of Wageningen UR.

“Civil society organisations are exerting pressure to make production more animal-friendly,” explains Tim Verwaart of LEI. “That leads to technological changes, such as the development of sustainable animal housing systems.” Other stakeholders that play a role in animal-friendly production are livestock farmers, abattoirs, supermarkets and consumers. The farmers’ scope for making investments and the choices consumers make are important, as is the way in which the supply chain is organised.

“The computer models, agent-based models as they are known, help us to understand the interactions between the various processes,” explains Verwaart. “That gives us a better understanding of the possible consequences of measures.” The ‘agents’ are pieces of software that represent individuals and that respond differently to changes in a system. This lets researchers study the behaviour both of individuals and of the systems within which they operate. A single simulation model can combine knowledge from e.g. economics, technology, ecology and social sciences. This yields better insights than an approach based on a single discipline.

According to Verwaart, “This is a young field that is still in its infancy, even worldwide.” Together with other divisions in Wageningen UR, we are exploring the possibilities in a research programme looking at ‘complex adaptive systems’. Models are being built for changes in land use, the spread of the potato blight Phytophthora and agrarian nature management. We are also working with the central veterinary institute CVI to develop a model for the spread of the hospital bacterium MRSA in pig farms. This model combines the development of resistance to antibiotics with the behaviour of livestock farmers and vets. “Interventions in complex systems such as these often have both foreseen and unforeseen consequences,” says Verwaart. “The model lets us work with other stakeholders to obtain a picture of the possible effects.”