Controls Engineer Intern

Foundation•San Francisco, CA

About The Position

We are building dexterous hands and fingers for an all-purpose humanoid — systems with 20+ degrees of freedom packaged in the volume of a human hand, that must move with precision, speed, and adaptability — and we need dedicated controls engineering bandwidth to make that happen Our hands are only as good as the control loops driving them; without tight, well-tuned control at the joint and finger level, even the best mechanism becomes unpredictable in contact with the real world We are a small team moving fast — this intern will not be shadowing anyone, they will own real subsystems, write firmware that runs on hardware, and contribute directly to milestones that matter The controls stack for a dexterous robotic hand is an unsolved problem at the frontier of robotics — impedance control, contact detection, tendon coupling, and real-time sensor fusion all need engineering attention that our current team cannot absorb alone We believe the best controls engineers are built early by working on hard hardware problems — this role is designed to give a talented student that foundation in a compressed, high-impact environment Every intern we bring in is a potential full-time hire; we are building the team that will ship the first truly capable humanoid hand, and we want people who want to grow with that mission

Requirements

Currently pursuing a BS or MS in Mechanical Engineering, Electrical Engineering, Robotics, or a closely related field — with coursework in control systems theory (state-space representation, Bode/Nyquist analysis, stability margins, PID design)
Foundational understanding of dynamics and kinematics for multi-body systems — you know what a Jacobian is, why it matters for robotic finger control, and how to use it
Exposure to modern control techniques beyond PID: state feedback, LQR, feedforward compensation, or impedance/admittance control — even if only from coursework or self-study
Understanding of sampled-data systems and discrete-time control — you know why sample rate matters for a fast motor control loop and what aliasing means in a sensor pipeline
Proficient in Python for simulation, prototyping, and data analysis — you can tune a controller in simulation before touching hardware
Working knowledge of C or C++ for embedded firmware — you are not afraid of pointers, interrupt handlers, or register-level peripheral configuration on microcontrollers (STM32, Teensy, ESP32, or similar)
Familiarity with at least one robotics framework — ROS2 preferred — including writing nodes, publishing sensor topics, and using standard tools like rqt and rviz for debugging