Promotion - End-to-End Trainable System for Autonomous Driving with Introspection

• Vollzeit
• Legal Entity: Robert Bosch GmbH

About the Company

At Bosch, we shape the future with high-quality technologies and services that inspire and improve people's lives. Our promise to our employees is unwavering: We grow together, take joy in our work, and inspire each other. Welcome to Bosch.

Robert Bosch GmbH looks forward to your application!

About the Position

The current state of the art in the field of automated driving (AD), represented by Vision-Language-Action models (VLA), allows vehicles to link perception and action more data-efficiently than classical end-to-end systems. However, VLA models are increasingly reaching their limits: due to a lack of long-term prediction capabilities, insufficient physical consistency, and limited coverage of long-tail scenarios. Therefore, another paradigm shift is necessary.

The goal of this research work is to investigate the next generation of AD systems and to shape the transition from reactive VLA models to predictive, agent-based world models. The central thesis is that "imagination before action" is an essential prerequisite for safe and scalable autonomy.

• As part of our team, you will develop a framework for integrating VLA models, world models, and agent-based planners into a holistic driving system.
• Furthermore, you will demonstrate through proactive simulation improvements in long-term predictions, handling of edge and special cases, and demonstrable safety.
• You will drive fundamental changes in data preparation and more efficient training of large systems by specifically considering unexpected scenarios.
• Lastly, you will contribute to the fundamental shift in the field of automated driving, away from reactive control towards simulation-based, agentic action, thereby paving the way for more powerful and trustworthy autonomous systems.

Responsibilities

• Develop a framework for integrating VLA models, world models, and agent-based planners.
• Demonstrate improvements in long-term predictions and handling of edge cases through proactive simulation.
• Drive changes in data preparation and training of large systems.
• Contribute to the shift from reactive to simulation-based, agentic action in autonomous driving.

Requirements

• Education: Excellent degree (Master or equivalent) in Computer Science, Robotics, Electrical Engineering, Mathematics, or a related field.
• Experience and Expertise: Solid knowledge of algorithms and principles of machine learning, preferably in the field of deep learning; proven programming skills in Python and C++.
• Personality and Work Ethic: You are an open-minded team player who approaches challenges with a goal-oriented, logical-analytical mindset.
• Languages: Fluent English in speech and writing, German skills are advantageous.

Additional Information

https://www.bosch-ai.com

www.bosch.com/research

The final assignment of the PhD topic is subject to your university as the central point of contact for your PhD.

Start: By agreement

Please send us your comprehensive application documents (including CV and certificates).

Diversity and inclusion are not trends for us, but firmly anchored in our corporate culture. Therefore, we are looking forward to all applications: regardless of gender, age, disability, religion, ethnic origin, or sexual identity.

Do you have questions about the application process?

Celina Vanessa Dannecker (Human Resources)

+49 711 811 21346

Do you have technical questions about the job?

Thomas Michalke (Technical Department)

+49 711 811 43435

Work #LikeABosch begins here: Apply now!

Job Location

• City: Stuttgart
• Country: Germany
• Remote: No

Apply Now

Jetzt bewerben

Data Protection Declaration

Imprint

Share this job