Postdoc Reinforcement Learning From Human Feedback Applied To Mobility System Design

TL;DR

Postdoc Reinforcement Learning from Human Feedback (Mobility System Design): Develop human-centred RL algorithms and recommender systems to guide stakeholders in urban mobility design with an accent on multiple objectives, robustness of designs, and transparency. Focus on RLHF techniques integrated with digital twins for transport, logistics, and liveable cities.

Location: Delft, Netherlands (relocation support to the Netherlands available)

Salary: €3546 - €5538

Company

Top international university combining science, engineering, and design to address global challenges in energy, climate, mobility, health, and digital society.

What you will do

• Develop AI-based recommender systems using RLHF for mobility system design in urban planning.
• Address algorithmic challenges like multiple objectives, design robustness, and recommendation transparency.
• Collaborate with Sequential Decision Making group, TNO, and Transport & Planning department on Mobility DesAIgn Lab project.
• Contribute to responsible use of digital twins in mobility, transport, and logistics.

Requirements

• PhD in AI, computer science, mathematics, or physics
• Thorough understanding of reinforcement learning
• Fluency in English

Nice to have

• Interest in multi-disciplinary projects involving transport engineering and societal partners
• Strong communication and teamwork skills
• Self-motivation for cutting-edge research

Culture & Benefits

• 2-year temporary contract, 36-40 hours per week
• Salary and benefits per Collective Labour Agreement for Dutch Universities
• Excellent pension scheme, flexible working week, 232 leave hours annually
• Opportunities for education, training, courses; partially paid parental leave; vitality program
• Diversity embraced; support for relocation including Coming to Delft Service and Dual Career Programme

Hiring process

• Apply by 21 May 2026 with CV and motivational letter
• Knowledge security risk assessment in final stages