Research Engineer - Vision Language Action Models for Intelligent Cyber Phy at Bosch Group

Renningen, Baden-Württemberg, Germany -

Full Time

Start Date

Immediate

Expiry Date

02 Jun, 26

Salary

0.0

Posted On

04 Mar, 26

Experience

5 year(s) or above

Remote Job

Yes

Telecommute

Yes

Sponsor Visa

Skills

Vision-Language-Action Models, Multimodal Representation Learning, Long-term Control, Planning, Reinforcement Learning, Imitation Learning, Model-based Control, Visual Grounding, Cross-modal Attention, Instruction-Following Architectures, Python, PyTorch, TensorFlow, JAX, Simulation Platforms, Scalable Pipelines

Industry

Software Development

Description

Company Description At Bosch, we shape the future by inventing high-quality technologies and services that spark enthusiasm and enrich people’s lives. Our promise to our associates is rock-solid: We grow together, we enjoy our work, and we inspire each other. Welcome to Bosch. The Robert Bosch GmbH is looking forward to your application! Job Description As a Research Engineer, you will develop cutting-edge Vision-Language-Action (VLA) architectures that empower AI agents to interpret human instructions and act autonomously in complex environments. Furthermore, you will connect multimodal representation learning with long-term control and planning to create agents that go beyond reactive capabilities and exhibit cognitive intelligence. You will contribute to fundamental and applied research on novel VLA models and actively drive their advancement. Building a scalable infrastructure for experimentation, training, and deployment, including the development of training pipelines, simulation tools, and evaluation methods will be part of your task. You will make VLA methods usable for concrete Bosch applications in practice and demonstrate their superior flexibility and generalization capabilities. Last but not least, you will work closely with interdisciplinary teams of researchers and application experts to shape the early innovation phase and the long-term strategy for intelligent automation at Bosch. Qualifications Education: excellent MSc in Computer Science, Machine Learning, Robotics or related technical fields PhD in Multimodal AI, Robotics, Reinforcement Learning, or Generative AI preferred demonstrated academic excellence, with a strong publication record in leading AI and robotics conferences and journals (NeurIPS, ICLR, ICML, CVPR, CoRL, RSS, ICRA, ACL, EMNLP, etc.) Experience and Knowledge: Industrial SW development experience: Demonstrated industry-relevant practical AI experience, e.g. by code contributions in industrial AI applications, in large scale machine learning projects, or participation in large-scale AI benchmarks and contests Multiple years of experience in developing and deploying machine learning solutions in distributed SW development teams Demonstrated capability to go beyond research prototypes and integrate cutting-edge AI methods into practically relevant and usable SW solutions Multimodal AI & Vision-Language Models: Proficiency in designing and training vision-language models (VLMs) and language-multimodal models (LMMs) (e.g., Flamingo, GPT-4V, PaLM-E, RT-2) Hands-on experience with visual grounding, cross-modal attention, and instruction-following architectures Familiarity with benchmarks such as ALFRED, Ego4D, VLN, or datasets involving perception and action Control, Reinforcement Learning & Planning: Strong experience in reinforcement learning, imitation learning, or model-based control for agents operating in real or simulated environments Capability in designing agents for long-horizon planning, semantic task decomposition, and hierarchical control Knowledge of methods that combine perception, language, and action in task-driven settings AI for Cyber-Physical Systems & Automation: Demonstrated capability to integrate cutting-edge AI methods into practical applications in robotics, automated driving, industrial automation, or building systems interest in emerging domains such as smart heating and HVAC control, where semantic AI can optimize control strategies Focus on building robust, explainable, and semantically grounded agents for physical deployment Infrastructure, Simulation & Tooling: Proficient in Python and deep learning libraries such as PyTorch, TensorFlow, or JAX Experience with simulation platforms such as Isaac Sim, CARLA, MuJoCo, or Habitat Skilled in developing scalable pipelines using Docker, CI/CD, and multi-GPU/cloud infrastructure Proven ability to drive research projects and turning research into practical innovation Collaborative and interdisciplinary mindset, able to work across research and product teams at Bosch Personality and Working Practice: you have proven ability to drive research projects and turning research into practical innovation; collaborative and interdisciplinary mindset, you are able to work across research and product teams at Bosch Languages: fluent English, German is optional Additional Information www.bosch.com/research https://www.bosch-ai.com Please submit all relevant documents (CV, certificates, and links to GitHub or kaggle account). We offer flexible working models: from various part-time options to mobile working and job sharing. Feel free to contact us. Diversity and inclusion are not just trends for us but are firmly anchored in our corporate culture. Therefore, we welcome all applications, regardless of gender, age, disability, religion, ethnic origin or sexual identity. Need support during your application? Meltem Arabacioglu (Human Resources) +49 174 1744961 Need further information about the job? Jim Mainprice (Functional Department) +49 711 811 21859 Michael Pfeiffer (Functional Department) +49 711 811 18195 Legal Entity: Robert Bosch GmbH

Responsibilities

The Research Engineer will develop cutting-edge Vision-Language-Action (VLA) architectures enabling AI agents to interpret instructions and act autonomously in complex environments, connecting multimodal learning with control and planning for cognitive intelligence. Responsibilities also include building scalable infrastructure for experimentation, training, and deployment, and applying VLA methods to concrete Bosch applications.