Here at Humanoid, we believe in a future where robots amplify human potential. That’s why we’ve set out on a mission to build the world’s most capable, commercially-scalable, and safe humanoid robots. We’re bringing that mission to life with HMND‑01 - our rapidly developed humanoid platform being deployed in real industrial environments - and we’re growing the team to take it even further.

About the Role

We are looking for a Senior Low-level Control Engineer to join our Control Team in London, focusing on control development and integration for our end-effectors.

You will design, implement and validate actuator- and joint-level controllers and grasp-level behaviours across a family of end-effector variants - parallel and multi-finger grippers, vacuum and suction tooling, and purpose-built tooling for specific customer tasks. Each variant brings different actuation, different sensing and, critically, different transmission mechanisms: geared direct drives, cable- and tendon-driven routings, differential and coupled linkages, and underactuated designs where few actuators must produce adaptive, compliant grasping across many degrees of freedom.

The ideal candidate has a solid background in control theory, strong software engineering skills in C++, and comfort working from real-time embedded control up through kinematics/dynamics and hardware debugging. You will work closely with mechanical and hardware engineers - early enough in the design cycle that your input shapes actuation and transmission choices rather than merely inheriting them.

What You'll Do

Low-Level Control Development

Tune and validate actuator- and joint-space control loops (PID, feedforward, impedance/admittance, observers, state feedback) for multi-DoF end-effectors in ROS2.

Develop grasp-level behaviours: grasp force control, compliant closing, slip detection and correction, and adaptive grasping on underactuated grippers.

Implement safety and fault-handling mechanisms — thermal and over-current protection, stall and jam detection, grip-loss detection, safe behaviour on fault.

Transmission Mechanisms, Modeling & System Identification

Build and identify models of actuators, transmissions and mechanisms - geared, cable/tendon-driven, differential, coupled and underactuated - including compliance, friction, backlash, hysteresis, cable stretch and pretension effects.

Own the joint-to-actuator abstraction for these mechanisms and the correct exposure of joint versus actuator state through the control stack.

Contribute directly to transmission mechanism design: work with mechanical engineers on routing, reduction ratios, DoF coupling and underactuation strategy, bringing the controllability and observability consequences of each option into the conversation early.

Sensor & Hardware Integration

Integrate and calibrate end-effector sensing (F/T, tactile, encoders, temperature/current, vacuum feedback) with real-time acquisition across the control stack.

Hands-on bring-up and debugging of new end-effectors, including test-stand work and vendor-supplied units.

Cross-Disciplinary Collaboration

Close the loop with mechanical/electronics engineers across sites on actuation, transmission and sensing requirements.

Work with data collection, applications and deployment teams so end-effectors work reliably in real use.

What We're Looking For

Shipped real-time-safe, object-oriented C++ control software running on real hardware.

Tuned and validated controllers on physical actuators, including sensor calibration and bring-up.

Modelled, identified and controlled through real mechanical transmissions - gearing, cables/tendons, differentials, coupled linkages or underactuated mechanisms - and dealt with the compliance, friction and backlash they introduce.

Worked on end-effectors or comparable multi-DoF mechanisms, including grasp control and supporting sensing.

Debugged at the fieldbus and OS layer: e.g. EtherCAT or CAN, on an RTOS or real-time Linux.

Strong control theory foundation (linear/nonlinear control, stability, observers, state estimation, feedback/feedforward, filtering, frequency-domain analysis) applied to multi-DoF mechanisms via Jacobians, numerical IK/FK, and URDF/Xacro.