Staff Robotics Control Engineer - Whole-Body Control (Wheeled Platform)
- Compensation
- Not disclosed
- Location
- UK, London
- Work setup
- Onsite
- Seniority
- Staff
- Employment
- Not specified
- Experience
- Not specified
- Category
- Industrial engineering
- Posted
Skills
Must have
- Deep, demonstrable expertise in optimisation-based whole-body control: hierarchical and weighted QP formulations, task prioritisation, constraint handling, and the numerical realities of solving them
- Strong command of the classical WBC toolkit - operational space and task-space control, inverse dynamics and IK formulations, redundancy resolution, nullspace projection, impedance and admittance cont
- Deployed whole-body controllers on real high-DoF hardware (humanoids, mobile manipulators, dual-arm or legged platforms), including coordinating base and manipulator motion.
- Practical fluency with QP/NLP solvers and their failure modes: conditioning, infeasibility, warm starting, solve-time budgets, and what to do at the limit.
- Strong dynamic modelling foundations: rigid body kinematics/dynamics (Jacobians, dynamics formulations, URDF-based modelling), inertial and friction system identification, and payload/external-load es
- Expert modern C++ for real-time control with Python for analysis and prototyping.
- Track record of taking control algorithms to production reliability rather than to a paper or a one-off demo, plus demonstrated technical leadership and mentoring.
Full description
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 Staff Robotics Control Engineer to join our Control Team in London and lead the evolution of whole-body control for our Wheeled Alpha platform - a humanoid whose whole body spans a wheeled base, torso and two arms, controlled as one coordinated system.
We need genuine depth in classical whole-body control. You should know hierarchical and weighted QP formulations from the inside - not as a library you have called, but as something you have derived, implemented, tuned, profiled and debugged on real hardware under demo pressure. You should have opinions about strict hierarchies versus soft-weighted stacks, regularisation and nullspace behaviour, what to do when a task set becomes infeasible, and how to keep a redundant system clear of singularities without it feeling sluggish. Alongside that depth, you will set technical direction and mentor a strong team of control engineers.
What You'll Do
Whole-Body Control Architecture & Algorithms:
- Lead the evolution of the Wheeled Alpha whole-body controller: task formulation, prioritisation, weighting, constraint handling and solver selection.
- Coordinate wheeled base, torso and dual-arm motion within a single whole-body formulation, with base motion that tracks commanded targets accurately and stays well-behaved under aggressive commands and long-duration operation.
- Own the constraint layer - self-collision avoidance, joint and velocity limit guarding, and singularity handling - plus feasibility handling and recovery when the solver fails.
- Drive the evolution toward predictive formulations: evaluate and develop MPC-based whole-body control where anticipating dynamics and constraints outperforms instantaneous QP resolution.
Dynamics, Compliance & Payload:
- Own the dynamic models the controller runs on - inertial, gravity and friction modelling, system identification, and online payload estimation - so the robot knows what it is carrying and controls accordingly.
- Advance compliance and force behaviour - impedance, admittance and force control that stays compliant at contact - and extend payload capability without sacrificing either.
- Own stability-aware behaviour under changing payload and dynamic motion