Agent Based Modelling for Social Simulation

This document is the result of an exploratory project looking into the status of, and opportunities for Agent Based Modelling (ABM) at TNO. The project focussed on ABM applications containing social interactions and human factors, which we termed ABM for social simulation (ABM4SS). During the course of this project two workshops were organized. At these workshops, a wide range of experts, both ABM experts and domain experts, worked on several potential applications of ABM. The results and insights from these workshops were used to produce this report, which contains guidelines related to the application of ABM and recommendations for its successful use within TNO. ---- The main conclusions of this project and report are the following: 1) Agent Based Modelling (ABM) is already successfully applied at TNO in numerous domains, but there is ample room for improvement. 2) TNO has a unique combination of knowledge on ABM and the various domains where ABM can be applied. However, this multi-disciplinary potential is currently not fully exploited. 3) ABM is a good candidate for many of the research questions which TNO faces. It is especially suitable for socio-technical systems. 4) The expertise within TNO is scattered across the organization. 5) The real challenge for TNO is the application of ABM to real world problems, not modelling platforms or technologies. This challenge consist primarily in determining if ABM is the proper approach to a specific problem and, if so, of collecting the required data to base the model on. ---- Although TNO has a unique mix of knowledge and expertise, there are a number of factors that are currently preventing TNO from taking full advantage of its potential. The following recommendations, that are explained in more detail in the report, are aimed at overcoming these limitations: 1) Focus research on development of real-world applications 2) Facilitate sharing of knowledge 3) Promote TNO’s knowledge position 4) Exploit multi-disciplinary potential 5) Consider alternative approaches 6) Investigate potential of big data and data-driven ABM 7) Investigate new areas of application. ---- The report concludes with a few outlines of ABM related projects that are proposed for the near future as a next step in the development of ABM as a research tool for TNO.