Robotics Control Engineer (Low-Level / Motor Control)

About the role

The AI Robot Association (AIRoA) is building a data platform for generative AI foundation models in robotics, collecting one million hours of humanoid robot operation data with more than 100 robots. This position designs, implements and tunes low-level control systems that turn motion commands from higher-level controllers into precise torque and motion for the joints of humanoid and dual-arm robots. It focuses on motor control (FOC, current/torque/velocity/position loops, board bring-up, sensor calibration, safety mechanisms) rather than motion planning or navigation.

Key responsibilities

• Design, implement and tune FOC (vector control) for brushless actuators, and current, torque, velocity and position control loops
• Bring up and debug custom motor-control boards and sensors
• Build calibration and synchronization pipelines for encoders and current/voltage measurements
• Implement actuator safety mechanisms such as overcurrent, overtemperature, watchdog and brownout protection
• Design control loops under real-time constraints using timers, ADC, PWM and DMA
• Implement CAN (FD) and EtherCAT interfaces and interfaces with higher-level software platforms
• Develop diagnostics and logging for torque, current, temperature and error states
• Collaborate with the Whole-Body Control and AI/ML teams on force and compliance control, and continuously tune control performance on hardware

Required profile

• At least 3 years of professional experience in real-time embedded software (microcontrollers, device drivers, low-level libraries)
• Strong C/C++ and practical motor-control knowledge (FOC, dq transformation, control loops)
• Experience with encoders, current/voltage sensors, SPI, UART, CAN (FD), RS-485 and EtherCAT
• Business-level Japanese; English technical reading and communication with overseas vendors

Required skills

• C/C++ and low-level peripherals (timers, ADC, PWM, DMA, non-volatile memory)
• Motor control: FOC, dq transformation, current/torque/velocity/position loops
• Preferred: RTOS, EtherCAT, CANopen, CiA 402, MATLAB/Simulink, MuJoCo, Isaac, Pinocchio, Python for testing