Robotics Controls Engineer

Humanoid; Boston (MA); On-site; USD 110,000 - 160,000; Full time

Job summary

Humanoid in Boston seeks a Controls Engineer to own testing and validation infrastructure for the robots. You will define metrics, build HIL workflows, and drive end-to-end validation on real hardware, enabling repeatable and measurable control performance.

Ideal candidates have strong control theory background, shipped C++ real-time software, and experience with test automation, observability, and hardware bring-up for complex robotic systems.

Qualifications

• Hands-on robotics experience on real hardware.
• Strong foundation in joint-space control, PID, and observers.
• Shipped software in C++ real-time stack and Python tooling.
• Built test automation and HIL/integration-test infrastructure.
• Proficient in measurement, statistics, and data analysis.
• Experience with observability tools such as logging and Grafana.

Responsibilities

• Build and own hardware-in-the-loop (HIL) and integration test workflows that quantify controller and motion repeatability and accuracy.
• Define and implement repeatability and accuracy metrics for controller and motion changes.
• Expand unit and integration test coverage across controls components: actuators, interfaces, whole-body control, teleoperation.
• Develop automated regression checks to surface controls regressions before wider release.
• Contribute to design of testability requirements for new components and algorithms.
• Package controllers into reusable, testable interfaces for isolation and interchangeability.
• Contribute to first-time bring-up of new hardware for Boston components and ensure end-to-end validation on real robots.
• Instrument the controls stack to improve visibility with logging and tracing.

Skills

Robotics experience

Control theory

C++ real-time

Python

HIL testing

Test automation

Observability tooling

Hardware bring-up

Debugging across HW/SW

Data analysis

Grafana

Tools

C++ RT stack

Python tooling

MuJoCo

Isaac Sim

Grafana

Job description

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 Alpha - our rapidly developed humanoid platform now running in real industrial pilots - and we're growing the team to take it even further.

We are looking for a Controls Engineer to join our Control Team in Boston.

You will own how we know our controls stack works: the testing and validation infrastructure that turns "it looked fine on the robot" into measured, reproducible evaluation, the observability and failure analysis that makes runtime behaviour visible and failures shareable, and the documentation and runbooks that let the team self-serve without institutional knowledge. You will also strongly contribute to first-time bring-up of new hardware for Boston components and validate changes end-to-end on real robots before the wider team depends on them.

What You'll Do

Testing & Validation Infrastructure:

Build and own hardware-in-the-loop (HIL) and integration test workflows that quantify controller and motion repeatability and accuracy - turning "it looked fine on the robot" into measured, reproducible evaluation.

Define and implement repeatability and accuracy metrics for controller and motion changes, and the harnesses that make those measurements consistent across runs and across robots.

Expand unit and integration test coverage across the controls components the team actually runs - actuator drivers, hardware interfaces, whole-body control, and the teleoperation pipeline.

Develop automated regression checks that surface controls regressions - in accuracy, repeatability, or behaviour - in simulation and on hardware before they reach the wider team.

Work with controls engineers to define testability requirements for new components and algorithms, so things are built to be evaluated and validated.

Package controllers into cleanly separable, reusable, and easily testable interfaces - so individual controllers can be developed, swapped, and validated in isolation.

New Hardware Bring-Up & Integration:

Strongly contribute to first-time bring-up of new hardware for Boston components, and to their integration into the controls stack.

Validate hardware and software changes end-to-end on real robots before broader team use - structured, documented, and repeatable.

Debugging, Observability & Failure Analysis:

Instrument the controls stack to improve visibility into runtime behaviour - logging, tracing, and diagnostic tooling.

Reproduce hardware and software failures in structured, developer-shareable formats; feed structured root-cause context back to algorithm and hardware owners.

Identify and resolve cross-stack failure modes that span controls, firmware, and hardware.

Documentation & Knowledge Transfer:

Write and maintain engineering documentation for the controls stack: HIL and test coverage, and bring-up/integration guides for new hardware.

Own runbooks that let controls engineers self-serve validation, integration, and recovery without relying on institutional knowledge.

What We're Looking For

Substantial hands-on robotics experience on real hardware.

Strong control theory foundation (joint-space control, PID and feedforward, impedance/admittance, observers, state estimation, filtering, stability) - enough to define meaningful repeatability and accuracy metrics for a control stack and interpret what they say.

Shipped software across the stack: modern C++ inside a real-time control codebase, plus Python for test harnesses, analysis and tooling.

Built test automation, HIL or integration-test infrastructure for complex physical systems - and got other engineers to actually use it.

Solid measurement and data analysis judgement: sampling, timing and synchronisation, noise, statistical significance, and knowing when a difference between two runs is real.

Strong debugging across software/firmware/hardware boundaries, with the discipline to reproduce failures in a form other engineers can act on.

Nice to have:

• Physics simulation environments (MuJoCo, Isaac Sim/Isaac Lab) and sim-to-real validation.
• Observability tooling: logging and tracing frameworks, metrics collection, time-series databases, dashboarding (e.g. Grafana).
• First-time hardware bring-up experience: new actuators, sensors or compute targets integrated into a control stack.
• A track record of engineering documentation and runbooks that outlived their author.

Release management or quality gates for safety-relevant sof