Control Engineer (m/f/d) Robotics

About the Job

We are looking for a Control Engineer (m/f/d) to bring locomotion control to our robot, from simulation to real hardware. You will own the full path from model to motor: identify the dynamics, design controllers, train reinforcement learning policies and make them run reliably on the real robot. The work spans simulation, real-time control and firmware. It has to hold up against model errors, sensor noise, latency and hardware that does not behave like the simulator.

Responsibilities

• Locomotion Control: Design, implement and tune controllers for a robot, from joint-level torque and impedance control up to whole-body and locomotion control.
• Real-Time & Firmware Integration: Integrate controllers and policies into the real-time control stack and work on the firmware side (motor drivers, communication buses, safety limits) together with the hardware team.
• RL Policy Training: Train locomotion policies in simulation (e.g. Isaac Lab, MuJoCo), including reward design, curricula and domain randomization.
• Sim2Real Transfer: Deploy trained policies on the real robot and close the sim2real gap through actuator, latency and noise modeling and systematic hardware experiments.
• System Identification: Identify actuator and rigid-body dynamics (inertia, friction, motor models) from robot data and feed the results back into simulation and controllers.
• Testing & Validation: Build test setups and benchmarks in simulation and on hardware, and run reliability checks before demos and releases.
• Documentation: Document models, controllers and procedures so others can reproduce and extend them.

Essential Skills

• Engineering Background: Degree in Robotics, Control Engineering, Electrical or Mechanical Engineering, Computer Science or a related field, or an equivalent qualification plus professional experience with comparable systems.
• Robot Dynamics & Control: Solid foundation in rigid-body dynamics and feedback control (e.g. torque/impedance control, MPC, whole-body control), applied on physical robots.
• Reinforcement Learning: Hands-on experience training RL policies for robot control in simulation (e.g. PPO with Isaac Lab or MuJoCo).
• Sim2Real: Experience transferring controllers or learned policies from simulation to real hardware and debugging the gap.
• C++ and Python: Professional experience writing C++ for real-time robot control on embedded targets, and Python for training, analysis and tooling.
• Hands-on Hardware Debugging: Comfortable working directly on real robots, reading logs and signals, and tracking problems across software, electronics and mechanics.

Beneficial Skills

• Legged Robots: Experience with quadrupeds or humanoids, in locomotion or whole-body control.
• System Identification: Experience identifying robot or actuator dynamics (excitation trajectories, physically consistent parameter estimation, friction models).
• Firmware & Embedded: Experience with embedded C/C++, motor control (FOC), real-time systems (e.g. PREEMPT_RT, Xenomai) or field buses such as EtherCAT or CAN.
• ROS 2 & Middleware: Experience with ROS 2, ros2_control or DDS.
• Simulation Modeling: Experience building or tuning robot models in MuJoCo or Isaac Sim.
• Infrastructure: Docker, CI pipelines and GPU training setups.