MS Thesis in collaboration with DENSO Automotive

Physical AI with VLA in Robotics

28.09.2026, Diplomarbeiten, Bachelor- und Masterarbeiten

Modern VLA models perform well on short‑horizon robotic tasks but often struggle with long‑term planning and multi‑step control. A promising direction is to enhance VLA architectures by integrating them with Physical AI models capable of understanding physical effects, e.g. for challenging, unstructured environments.

For such VLA-based physical AI systems, the project aims to evaluate and quantify a robot’s action‑generation capability of a physical AI VLA model for more challenging, unstructured environments, and to develop a clearer understanding how these contribute to more stable and capable long‑horizon action generation.

Research Areas

• Survey state‑of‑the‑art VLA models and world‑model approaches

• Implementing and benchmarking physical variants with a VLA framework

• Developing simulation‑based setups for model evaluation for challenging scenarios.

• Exploring and validating improved integration strategies based on findings (if feasible)

• Identifying and designing task‑specific strategies for fine-tuning to enhance model performance (if feasible)

Prerequisites:

• Basic understanding and strong interest in Robotics

• Basic understanding of computer vision

• Ability to understand and run modern deep learning and robotics codebases

• Proficiency in Python or C/C++

• Experience in physical AI, world models, and simulation a plus.

Kontakt: christian.prehofer@tum.de