Simulation Engineer – Cognitive Twin (human)

About NEURA Robotics and the Cognitive Twin

At NEURA Robotics, we create cognitive robots and develop Neuraverse, an integrated ecosystem that connects robots, their skills, and the developers behind them. The Cognitive Twin serves as the comprehensive platform where robotic applications are designed, real-time data visualized, and simulations run on both virtual and physical robots. This allows applications to be developed and verified in simulation well before deployment on actual hardware.

The future of robotics involves establishing software validity through simulation instead of hardware first. On the Cognitive Twin team, you will bring the simulation features to life by integrating robots, sensors, and physics scenarios, enabling teams to simulate, benchmark, and validate performance against real-world conditions.

Key Responsibilities

• Enable and operate simulation capabilities within the Cognitive Twin environment, utilizing the platform's physics engine to configure rigid-body dynamics, contacts, and articulations, and integrating robot, sensor, and environment models for consistent, repeatable scene execution.
• Construct and parameterize physics simulation scenes, including headless setups; connect sensors and robots and expose simulation functions accessible through web interfaces and APIs.
• Develop standardized headless benchmark scenes and test cases; automate in-loop executions to assess control, planning, and learning techniques against fidelity and performance parameters such as step rate, parallel instance count, latency, and memory usage.
• Perform simulation to real-world comparisons by analyzing twin behavior against logs from physical robots; identify simulation-real world discrepancies like dynamics or sensor noise differences and assist in calibrating simulation parameters as fidelity requirements evolve.
• Create and maintain pipelines for importing and exporting common 3D and robot description formats ensuring high fidelity and smooth interoperability for scenes entering or leaving the twin system.
• Manage simulation backend concerning scene definition formats: map composed scenes to the engine's runtime representation and oversee state output via real-time binary streams in collaboration with Integration, Frontend, and AI/ML teams.
• Maintain reproducibility and quality by version controlling scenarios, configurations, and results; establish modeling conventions and acceptance criteria; and generate structured reports documenting benchmark outcomes.

Qualifications and Skills

• A bachelor's or master's degree in Robotics, Mechanical or Electrical Engineering, Computer Science, Mechatronics, or related disciplines, or equivalent practical experience.
• Multiple years of relevant professional experience in robotics simulation, benchmarking, or model-based assessment within academia or industry.
• Practical experience with modern real-time 3D or game-engine-based simulation platforms and production physics engines such as PhysX, involving configuration of rigid-body dynamics, articulations, and contact modeling; proficient in robot and environment modeling with common robot-description and scene interchange standards.
• Solid theoretical and practical knowledge of multi-body dynamics, kinematics, and contact physics to confidently configure and validate simulation models.
• Strong proficiency in modern C++ (C++17) and CMake, complemented by Python skills for tooling; capable of navigating complex, multi-language codebases.
• Competence in defining quantifiable metrics and statistically analyzing simulation outcomes, alongside experience in structured experiment management and automation of simulation testing workflows including batch runs and CI-like testing.
• Familiarity or keen interest in sim-to-real transfer methodologies and well-informed about real-world robotics hardware and its constraints.
• Experience with AI-assisted coding tools such as Claude Code, Cursor, Copilot, and familiarity with agent-based development orchestration.
• Additional desirable skills include experience with MuJoCo, NVIDIA Isaac Sim, PyBullet; GPU-accelerated or distributed simulation techniques; modeling of sensors and actuators including IMUs, LiDAR, cameras, and force-torque sensors; and expertise in scene interchange tooling.
• A product-focused mindset emphasizing delivering user value and adaptability to changing product requirements.
• Organized work habits, team-oriented approach, and precision in technical communication; fluent English required, German language skills advantageous.