PhD position in Agent-based Modelling for Inspection Simulations

Join our exciting research project as PhD candidate at Utrecht University and become part of the research collaboration AI4Oversight Lab which is part of the Innovation Centre for Artificial Intelligence.

Your job
In this PhD position, you will focus on simulating the inspection processes. Strategies that inspectorates use for selection of targets and enforcement in response to non-compliances invoke changes in the behaviour of the target population. In turn, these changes invoke updates in the strategies of the inspectors. It is difficult—if not impossible—to study and test the interactions between an inspectorates’ actions and the behaviour of inspectees outside of a lab setting because of practical and ethical constraints. In this project we aim to develop data-driven agent-based models to study these behavioural dynamics in a synthetic environment, and thereby enabling the evaluation of (data-driven) inspection strategies before field deployment, while testing for effectiveness, robustness, and fairness of target selection strategies.

The Dutch government inspectorates play a critical role in safeguarding public interests such as food safety, a clean environment, fair working conditions, and quality of education. To ensure effective supervision with a limited capacity at strategic and operational level, inspectorates need to work in a data-driven way and embed AI technology in their primary processes. The amount of information is too large and complex to be fully covered by human resources. At the same time, society rightly expects AI to be used in a responsible manner.

By joining the AI4Oversight ICAI lab, you join a collaborative community that addresses AI challenges specific to the inspection domain leading to scientifically assured methods. The AI4Oversight lab connects the Human Environment and Transport Inspectorate (ILT), the Netherlands Labour Authority (NLA), the Inspectorate of Education (IvhO), Netherlands Food and Consumer Product Safety Authority (NVWA), Netherlands Organisation for Applied Scientific Research (TNO), Utrecht University and Leiden University. Collaboration between these organisations is seen as an essential element of our lab. Working together allows not only to develop new knowledge, but also to use each other’s expertise, to experiment together, to learn from each other and to bring theory to practice.

The execution of the research will be highly participatory. You will spend time at the offices of funding partners and have the opportunity to dive into the practical challenges and way of working of the partners. You will work together with data scientists of the practical partners, who will contribute with practical experiences and use cases. The AI4Oversight Lab hosts at least 6 PhD candidates and knowledge exchanges will be held regularly to promote collaboration.

Key responsibilities:

• conducting original and novel research in the field of data-driven simulations and large-scale agent-based modelling in the domain of government oversight;
• publishing and presenting scientific articles at international journals and conferences;
• collaborating with other PhD candidates in the AI4Overisight lab, researchers at the partners’ data science labs, the intended users, and other stakeholders.