JOB DESCRIPTION

We invite applications for a postdoctoral researcher position at Delft University of Technology (TU Delft) focused on AI-supported traffic prediction and planning, spanning both short-term operational decisions and long-term infrastructure interventions. You will contribute to two integrated use cases within the AIM-TT (AI Methods for Traffic and Transportation) program, in collaboration with Rijkswaterstaat, NDW, TU Eindhoven, Arane, and d-fine.
Use Case 1 focuses on the real-time prediction of traffic flow and capacity under acute disruptions—such as incidents, roadworks, or active traffic management—on a sensitive motorway segment near the Ketheltunnel (A4). Your task will be to develop hybrid AI models that fuse traffic engineering principles with machine learning to provide realistic, explainable, and timely predictions. These models will support operators in selecting appropriate traffic management measures under uncertainty.
Use Case 2 explores the use of AI for the planning and phasing of large-scale roadworks, using the reconstruction of the A13–A16 connection near Rotterdam as a real-world testing ground. Your role involves identifying causal relationships between interventions and traffic outcomes, developing prediction models, and analyzing “what-if” scenarios to support multi-objective planning decisions.
You will work closely with a second postdoc at TU Eindhoven who focuses on simulation and optimization, while your focus lies on prediction, data fusion, and causal modeling. The postdoc position is embedded in the Transport & Planning section of the Faculty of Civil Engineering and Geosciences, offering access to a dynamic research environment and a strong international network.

Your responsibilities include:

• Developing hybrid AI models for traffic prediction under disrupted and planned conditions.
• Fusing real-time and historical data from various sources (e.g. loop detectors, FCD).
• Collaborating with TU Eindhoven on the development of a joint decision support and simulation environment.
• Publishing in high-impact scientific journals and conferences.
• Supporting MSc thesis supervision and contributing to research proposals and outreach activities.

The project offers a unique opportunity to work on real-world, high-stakes challenges in mobility with direct societal relevance, combining academic depth with practical applicability.

JOB REQUIREMENTS

We seek a motivated and collaborative researcher who brings both technical expertise and an interest in applied, societally relevant AI research. The ideal candidate has:

• A PhD (or is close to completion) in Transportation Engineering, Artificial Intelligence, Data Science, or a related field.
• Demonstrated experience in traffic modeling, traffic operations, or behavioral modeling.
• Familiarity with AI methods such as time-series modeling, Kalman filtering, causal inference, or deep learning.
• Strong programming skills (e.g., Python, MATLAB) and experience with mobility datasets (e.g., loop detectors, floating car data).
• A strong interest in hybrid approaches combining machine learning with domain knowledge.
• Excellent communication skills in English and an ability to work in an interdisciplinary team.

Experience with real-time traffic prediction, scenario analysis, or traffic planning tools is considered a plus.

• Developing hybrid AI models for traffic prediction under disrupted and planned conditions.
• Fusing real-time and historical data from various sources (e.g. loop detectors, FCD).
• Collaborating with TU Eindhoven on the development of a joint decision support and simulation environment.
• Publishing in high-impact scientific journals and conferences.
• Supporting MSc thesis supervision and contributing to research proposals and outreach activities