Robot Simulation & Synthetic Data Generation Working Student / Intern (m/f/d)

YOUR TASKS

• Collaborate with robotics and AI engineers to integrate NVIDIA Isaac Sim as a simulation environment for a heterogenous fleet of mobile robots.
• Generate synthetic multimodal training data for SOTA AI models in the field of robotics using NVIDIA Isaac Sim.
• Create realistic simulation environments for robots based on real sites using digital twin data and data collected by our robots in the field.
• Assist in designing and refining simulation scenarios that replicate real-world robotics use cases.
• Contribute to the creation, validation, and organization of synthetic datasets tailored for machine learning and computer vision tasks.
• Help extend internal documentation and contribute to team knowledge sharing on simulation best practices and reproducible synthetic data generation pipelines.
• Participate in interdisciplinary team meetings and research sessions to enhance simulation realism and data quality.
• Document your workflows, results, and findings in a clear and accessible manner.

YOUR PROFILE

• Currently enrolled in a Bachelor’s or Master’s degree program in Robotics, Computer Science, Data Science, Engineering, or a related technical field.
• Experience with robotics simulation or synthetic data generation environments such as NVIDIA Isaac Sim, Gazebo, Carla, Unity or Blender.
• Experience with ROS/ROS2.
• Experience in Python programming and familiarity with scientific computing libraries (NumPy, Pandas, etc.).
• Experience with or existing knowledge of ML/AI in robotics and computer vision.
• Good grasp of data concepts: dataset generation, cleaning, and annotation.
• Working knowledge of machine learning and computer vision principles.
• Strong verbal and written communication skills in English.

HELPFUL ADDITIONAL QUALIFICATIONS

Nice to have: 

• Experience with C/C++ programming, preferably in performance-critical applications.
• Experience with 3D modelling/CAD programs such as Blender, Fusion 360, FreeCAD and OpenSCAD.
• Experience with state-of-the-art generative models for 2D/3D data.
• Experience with annotating or converting datasets for standard ML formats (COCO, KITTI, etc.).
• Exposure to or interest in LiDAR sensor simulation and 3D perception.
• Experience with image-based 3D reconstruction approaches such as SfM, Gaussian Splatting, NeRFs.
• Experience with Docker.
• Previous contributions to open-source simulation or robotics projects.