Ministry of Transport – Developing Assumptions for a National Agent-Based Model

The goal was to produce a highly detailed, anonymised dataset of freight movements to underpin agent-based simulations, giving the Ministry an evidence-based foundation for scenario testing and long-term infrastructure planning.

Our Approach:

  • OD Matrix Development – Processed GPS freight trip data to create an SA2-level Origin–Destination matrix, showing the intensity and direction of freight movements between regions.
  • Data Privacy Protections – Implemented strict anonymisation measures, ensuring that no customer home-base locations could be identified from the dataset.
  • Integration with Agent-Based Model – Prepared and delivered the OD matrix in a format compatible with the Ministry’s national agent-based model, enabling immediate scenario testing.

Key Findings:

  • Granular Freight Flow Insights – Provided a clear, geographically detailed picture of freight movements across New Zealand.
  • Scenario-Driven Planning – Enabled the Ministry to test the effects of various future scenarios, including economic growth, infrastructure upgrades, and policy shifts.
  • Strategic Investment Signals – Highlighted priority areas for infrastructure development to maximise freight efficiency and network resilience.

Impact:

  • Evidence-Based Policy Development – Grounded policy discussions in real-world freight movement data, increasing confidence in planning decisions.
  • Future-Ready Investment Planning – Informed long-term infrastructure priorities by identifying freight corridors most sensitive to change.
  • Balanced Impact Assessment – Allowed the Ministry to weigh economic, social, and environmental factors when assessing potential policy interventions.

Why Work With Robinsight

Privacy First

We align with New Zealand’s Data Protection and Use Policy and data.govt.nz stewardship standards to protect privacy at every step.

Bespoke Analysis

We start with your specific transport question, then design data pipelines and models to answer it precisely.

Proven Methods

From spatial–temporal matching to machine learning, we apply the right tools to deliver regulator-aligned, decision-ready outputs.