From Research to Real-World Traffic Management

DeepTraffic is not developed in isolation. The research is closely connected to real-world traffic management challenges and is tested in two use cases.

These use cases provide realistic environments in which the models can be developed, tested and evaluated together with practitioners.

Use case #1: Predictive traffic management at the Ketheltunnel

The Ketheltunnel on the A4 between Delft and Rotterdam is a critical location where congestion needs to be prevented from developing inside the tunnel.

DeepTraffic investigates how hybrid traffic prediction can support real-time operational decisions when conditions change rapidly — for example because of incidents, temporary closures or traffic management interventions.

The research combines multiple sources of traffic information with domain knowledge to produce predictions that are both reliable and interpretable.

Use case #2: Planning large-scale roadworks around the A16

Large infrastructure projects can fundamentally change traffic demand and route choice. DeepTraffic investigates how predictive models can support the planning and phasing of major roadworks around the A16 corridor near Rotterdam.

By combining traffic data, causal modelling and simulation, the project explores how different planning and detour scenarios may affect traffic flows and network performance.

The aim is to help identify strategies that minimise disruption and societal costs.