Simulation Engineer – Cognitive Twin (Mensch)

The role

You will build and operate simulation capabilities for the Cognitive Twin based on the platform's physics engine. You will configure robot, sensor, and environment models to enable teams to simulate, benchmark, and validate against the real world.

What you will do

• Build and operate simulation capabilities based on the platform's physics engine
• Configure rigid-body dynamics, contacts, and articulations
• Build and parameterize physics scenes and headless setups
• Make simulation features accessible via browser and API
• Build standardized headless benchmark scenes and test cases
• Evaluate control, planning, and learning approaches using fidelity and performance metrics
• Compare Twin behavior with real robot logs to identify sim-to-real gaps
• Build and maintain ingest and export pipelines for 3D and robot description formats
• Implement backend simulation scene definition and state output
• Maintain versioned scenarios, configurations, and results

Required qualifications

• Bachelor or Master degree in Robotics, Mechanical Engineering/Electrical Engineering, Computer Science, Mechatronics, or a related field, or equivalent practical experience
• Multiple years of relevant experience in robotics simulation, benchmarking, or model-based evaluation in research or industry
• Practical expertise with a modern, real-time 3D/game-engine-based simulation environment and a production-ready physics engine (e.g. PhysX)
• Experience with robot/environment modeling using common robot-description and scene-interchange formats
• Solid understanding of multi-body dynamics, kinematics, and contact physics
• Strong modern C++ (C++17) and CMake
• Python for tooling
• Experience with large, multilingual codebases
• Experience defining objective metrics and statistical evaluation of simulation results
• Experience with structured experiment management and automation of simulation tests
• Experience with, or strong interest in, sim-to-real transfer and understanding of real robot hardware
• Familiarity with AI coding tools (e.g. Claude Code, Cursor, Copilot) and agent-based orchestration

Nice to have

• MuJoCo
• NVIDIA Isaac Sim
• PyBullet
• GPU-accelerated/distributed simulation
• Actuator and sensor modeling (IMUs, LiDAR, cameras, force-torque sensors)
• Tooling for scene-interchange formats
• German

About the company

NEURA Robotics builds cognitive robots and the Neuraverse, a networked ecosystem where robots, skills, and developers come together. The Cognitive Twin is a unified environment for designing, visualizing, and simulating robotic applications.