Controls Engineer - Actuation

Apptronik•Austin, TX

2d•Onsite

About The Position

Apptronik is building robots for the real world — to improve human quality of life and help solve the ever-increasing labor shortage. As a Controls Engineer – Actuation on our Controls team, you will own both the actuator models that make our simulation trustworthy and the actuator-level control that runs on real hardware — closing the loop from bench characterization and system identification, through simulator integration, to controller design, tuning, and deployment. You will pair deep motor electromechanics and system-identification skill with strong C++/Python engineering, working hand in hand with our Controls, Firmware, and Simulation teams so that both model fidelity and control performance are driven by how they are actually used on the robot.

Requirements

Strong C++, including comfort extending a simulator or physics-engine codebase — not just consuming one through a Python interface.
Deep grounding in motor electromechanics and classical control — closed-loop torque/current control, controller design and tuning, and frequency-domain analysis (e.g. Bode plots, stability margins).
Hands-on actuator or servo control on real hardware: designing, tuning, and validating low-level controllers, informed by system identification.
Hands-on system identification and modeling: structured experiment design and model fitting from real data that feeds both simulation and control.
Proficiency in Python (test automation, data analysis, tooling) and a track record shipping maintainable code under Git in a fast-moving R&D environment.
Comfortable working cross-functionally with controls engineers — focused on how your models and controllers get used, not just whether they fit the data.

Nice To Haves

experience with dynamometers, motor test stands, and lab instrumentation
motor thermal modeling and gearbox efficiency / backlash characterization
real-time actuator communication protocols and embedded / firmware control
a modern robotics simulator (MuJoCo, Isaac Sim/Lab, Brax) and sim-to-real validation workflows
Docker with GPU/cloud experiment infrastructure
Prior humanoid or legged-robot experience is strongly preferred.

Responsibilities

Develop and maintain high-fidelity actuator models in simulation, capturing the physical behaviors (friction, compliance, thermal, torque and current limits) that meaningfully affect real-world performance.
Design, tune, and deploy actuator-level control — closed-loop torque and current control and the low-level control that turns commands into reliable joint behavior on hardware.
Design and run hardware test campaigns to characterize actuator behavior and extract the parameters that keep both the models and the controllers accurate.
Apply system identification to extract trustworthy physical models from experimental data, and use those models to inform controller design and gain selection.
Quantitatively validate sim-to-real fidelity — for both the models and the controllers — using appropriate comparison metrics, and understand the limitations of each.
Build and maintain robust Python/C++ tooling for experiment automation, model extraction, controller tuning, and simulator integration.
Partner with the Controls team throughout — aligning on the fidelity and control performance actually needed, how model and controller choices ripple into training and policy performance, and where the actuator layer is the root cause of a sim-to-real gap.
Support hardware bring-up and deployment, translating validated models and controllers into real robot configuration.