Master Thesis Student (m/f/d) - Physical AI with Vision Language Action (VLA) Models in Robotics

DENSO Europe; Eching am Ammersee; Ibrido; EUR 13.000 - 20.000; Tempo pieno

Vantaggi offerti da questo lavoro

Home office

Health & wellbeing days

Urban Sports Club

Canteen onsite

Free parking

Team events

Descrizione del lavoro

DENSO Europe is seeking a Master student in CS/Engineering to explore and evaluate Vision-Language-Action models integrated with physical AI for long-horizon robotic tasks. You will work on understanding how action generation works in unstructured environments and help quantify improvements.

This role involves surveying current VLA and world-model approaches, implementing and benchmarking variants, building simulation setups, and proposing task-specific fine-tuning strategies.

Competenze

• Enrolled in a Computer Science/Engineering Master with focus on data analysis or robotics.
• Strong interest in robotics and basic understanding of computer vision.
• Ability to understand and run modern deep-learning and robotics codebases.
• Proficiency in Python or C/C++.
• Experience with physical AI, world models, and simulation is a plus.

Mansioni

• Survey state-of-the-art VLA models and world-model approaches.
• Implement and benchmark physical variants using a VLA framework.
• Develop simulation-based setups for model evaluation in challenging scenarios.
• Explore and validate improved integration strategies based on the findings.
• Identify and design task-specific fine-tuning strategies to improve model performance.

Conoscenze

Robotics basics

Computer vision basics

Python

C/C++

Deep learning codebases

Data analysis interest

Formazione

Master's student in CS/Engineering

Strumenti

Simulation

World models

Descrizione del lavoro

Company Description

As an automotive supplier, DENSO is leading in developing and providing components and systems for heating, air conditioning, motor cooling, exhaust gas aftertreatment, automotive electrics and electronics and instrumentation.

Job Description

Modern Vision Language Action (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;


• Implement and benchmark physical variants using a VLA framework;


• Develop simulation-based setups for model evaluation in challenging scenarios;


• Explore and validate improved integration strategies based on the findings (if feasible);


• Identify and design task‑specific fine‑tuning strategies to improve model performance (if feasible).

Qualifications

• Currently enrolled in a Computer Science / Engineering Master studies with a focus on data analysis, automotive, engineering or similar


• Basic understanding of 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 with physical AI, world models, and simulation is a plus

Additional Information

• Home office: Possibility to work from home


• Health: Yearly health & wellbeing days, Urban Sports Club membership


• Attractive working environment: Modern offices, canteen onsite and free parking


• Open company culture: Diverse, multicultural team


• Events: Regular company & team events