Agent-based modelling of regional innovation systems: A literature review

Since the early 1990s, our understanding of how innovations develop has changed a lot. We observe a clear shift in the modeling of innovation systems in the literature. Traditionally, knowledge inputs were somehow directly linked to outputs. However, this linear (and location-based) view soon proved to be insufficient as it neglected the importance of the underlying microstructure of regions and the complexity of regional knowledge creation processes. Modeling regional innovation systems is a very challenging task, as various influencing factors have to be taken into account. In this respect, empirically based agent-based models have proven to be a promising approach. With advances in computing power and the availability of individualized data, agent-based models can model real-world scenarios and provide deeper insights into regional innovation systems. This article provides a historical background to the systems perspective of regional innovation and a comprehensive summary of the latest work on agent-based modeling as applied innovation systems, with a specific focus to the regional level. The aim is to explore the various methodologies, frameworks, and case studies and enhance our understanding of regional innovation dynamics and agent-based modeling. Despite their potential, much of the work is still purely theoretical, lacking an empirical basis and therefore unable to address real-world problems. The main challenges for current and future research are therefore the development of a consistent validation strategy that allows the replication of the models. We also argue that there are still only a limited number of studies that analyze regional innovation systems and include elements of participatory foresight in their agent-based models, although it has the potential to advance the scientific debate in the field of innovation geography.