Senior Software Controls Engineer

About the role

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 Software Engineer to join our Control Team in London.

You will own how we measure control quality - per robot and across the fleet. Today, judging whether a change made the robot better is still too often an engineer's opinion after a few runs. You will replace that with numbers: reproducible, automated, trusted numbers that say whether a release is better than the last one, whether one robot behaves like another, and whether a fleet is ready for a pilot.

This is a software engineering role, not a control design role. You do not need to derive a whole-body controller. You do need real robotics experience, strong engineering discipline, and genuine interest in the physics of what you are measuring - the control engineers bring the theory and will rely on you to tell them the truth about their code.

WHAT YOU'LL DO

Control Quality Metrics

- Turn qualitative concerns about control quality - smoothness, latency, tracking error, accuracy, and the like - into precise, computable metrics, and build the pipelines that compute them consistently from recorded robot data, simulation runs and live telemetry.

- Own metric definitions and versioning so a number measured today is comparable with the same number measured six months from now.

Evaluation & Test Infrastructure

- Develop automated test harnesses that evaluate control releases deterministically - reproducible, run-to-run comparable, free of the hidden state that makes hardware testing unreliable.

- Wire control evaluation into CI across simulation and hardware-in-the-loop, with pass/fail thresholds so regressions are caught automatically rather than discovered on a customer site.

- Build the tooling that lets engineers compare controller variants and releases side by side.

Fleet-Level Control Quality

- Build the per-robot and per-fleet test suites that establish whether each robot meets the expected control performance bar - runnable by other teams without controls support.

- Track control quality over time and across units - drift, per-robot variation, degradation - and own the path from robot to metric, including the dashboards that make control health legible.

WHAT WE'RE LOOKING FOR

- Robotics experience on real systems - you understand robot data, timing and synchronisation, and the many ways hardware experiments go wrong.

- Strong Python, plus enough modern C++ to work confidently inside a real-time control codebase and instrument it.