[Jobs] PhD & Postdoc Positions in Structured Robot Learning & Generative AI for Robotics – TU Darmstadt (PEARL lab)

[Georgia Chalvatzaki]

Dear colleagues,

We are seeking highly motivated PhD Students and Postdoctoral Researchers to join the PEARL Lab (https://pearl-lab.com) at TU Darmstadt, Germany, led by Prof. Georgia Chalvatzaki.

Our Research Vision

Our research focuses on structured robot learning and embodied AI. We aim to enable robots to perceive, reason, and act in open-ended environments by leveraging the underlying structure of robot–environment systems. Our approach moves beyond black-box policies toward composable, adaptive, and robust behaviors that generalize across tasks, supporting grounded, language-guided decision-making and intuitive human-robot interaction.

Key Research Directions

We are particularly looking for candidates interested in:

• Grounding for Embodied AI: Connecting language and high-level plans to structured perception, affordances, and executable actions.
• Grounded World Models: Learning structured representations of geometry, semantics, and dynamics for prediction and interaction.
• Foundation Models for Robotics: Multimodal models (VLA) for long-horizon decision-making, interfacing with structured control modules, memory, and composable low-level skills.
• Reinforcement Learning: Structure-aware policy learning with efficient long-horizon credit assignment and RL-based fine-tuning for offline-to-online adaptation.
• Mobile Manipulation: Embodied decision-making for whole-body, contact-rich interaction and bimanual coordination in dynamic environments.

Research Opportunities

The successful candidates will contribute to two major, well-funded research initiatives:

• European AI Lighthouse Project: Join the brand-new EU project on Generative AI for Robotics — a flagship "Lighthouse" initiative developing the first European Foundation Models for Robotics made in Europe.
• ERC Starting Grant SIREN: Be embedded within the project SIREN (Structured Interactive Perception and Learning), researching the principles of robotic embodied intelligence through a holistic view of robot–environment interaction.

Environment & Collaboration

TU Darmstadt is a leading European institution for AI and Robotics. You will be part of an exceptional ecosystem, including:

• The Hessian.AI center for artificial intelligence.
• The Clusters of Excellence Reasonable Artificial Intelligence (RAI) and The Adaptive Mind (TAM).
• State-of-the-art laboratory facilities with advanced humanoids, mobile manipulators, and robot arms.

Qualifications

We are looking for candidates with a strong background in Robot Learning, Robotics, Learning for Control, Computer Vision, Machine Learning or related areas.

• Technical Skills: Proficiency in Python/PyTorch is expected.
• Experience: Prior work with robot learning, RL, foundation models, or real robot hardware is a strong plus.
• Postdocs: Should demonstrate a strong publication record and the ability to lead independent research.

How to Apply

Please send your application through our website https://pearl-lab.com/open-positions including:

• A detailed CV.
• Academic transcripts (for PhD applicants).
• A brief motivation letter.
• A brief research statement.
• Representative publications or project links.

Informal inquiries are very welcome!

Best regards,

Georgia Chalvatzaki

Professor for Interactive Robot Perception and Learning (PEARL),
Intelligent Robotic Systems for Assistance (iROSA) Emmy Noether DFG group,
https://pearl-lab.com

Computer Science Department
Technische Universität Darmstadt 
https://www.tu-darmstadt.de
Hessian.AI