Overview

⏱️Schedule:  Doors 18:30 → Talk 19:00→ Q&A → Mingle
🌐Language: English

Criminal networks don't want to be studied. Every transaction is hidden, every relationship deniable — so how do you build a model of something designed to stay invisible?

Frederike Oetker has tackled this problem twice, in two very different ways. For her PhD, working directly with the Dutch police, she built a Digital Twin of the Netherlands' cocaine trade — a full computational replica of the real network, built on rich qualitative and quantitative data most researchers simply don't have. With it, she could answer questions such as: what happens to a network when a kingpin is imprisoned, and who replaces him?

Now, as a postdoc on the FORT-PORT project, she's facing the opposite problem: modeling cocaine trafficking through the Port of Rotterdam using open, public data with huge gaps. Her solution is to impute the missing pieces and build a complete econometric model — one that can forecast how much cocaine is likely to move through the Netherlands in the coming years, which routes traffickers are likely to take, which countries might see import spikes if the Port of Rotterdam were closed off, and where hidden drug labs might be operating.

What we'll cover

• How rich police data made it possible to build a Digital Twin of the Dutch cocaine trade — and what it revealed about kingpin replacement in criminal networks
• The opposite problem: modeling trafficking through the Port of Rotterdam using incomplete, public data
• How missing data gets filled in — and why the resulting predictions can still be trusted
• Forecasting the future: expected trafficking volumes, likely smuggling routes, which countries might feel the effects of port closures, and where hidden drug labs might be operating in the Netherlands

Speaker

Frederike Oetker

Postdoctoral Researcher, Applied Probability Group — TU Delft

Frederike Oetker is a postdoctoral researcher in TU Delft's Applied Probability Group, working on the interdisciplinary FORT-PORT project on cocaine trafficking through the Port of Rotterdam. She did her PhD on cocaine networks in the Netherlands at the University of Amsterdam's Computational Science Lab, where she built a Digital Twin of the Dutch cocaine trade in collaboration with the police. Her work spans agent-based modeling, data imputation, machine learning, and causal loop diagrams — using clandestine networks as her subject.