Controls Lead
Controls Lead
- Location San Francisco
- Expertise Robotics
- Job Type
- Salary $ 270,000 per annum
Controls Lead
About the company
Our client is a well funded early stage robotics company building a standalone, battery powered, multi legged robot for critical infrastructure, with utilities as the initial focus. They are raising a funding round and are hiring their first tranche of engineers to build out the software and systems side of a hardware strong team.
The role
The Controls Lead owns locomotion and whole body control for a multi legged robot working on uneven terrain. Model based control is the priority, and learning based control is a bonus, with a research team covering reinforcement learning. This is a hands on role with no direct reports to begin with, reporting to the VP Engineering, with a path to build a controls team as the company scales.
Key responsibilities
Apply Now
About the company
Our client is a well funded early stage robotics company building a standalone, battery powered, multi legged robot for critical infrastructure, with utilities as the initial focus. They are raising a funding round and are hiring their first tranche of engineers to build out the software and systems side of a hardware strong team.
The role
The Controls Lead owns locomotion and whole body control for a multi legged robot working on uneven terrain. Model based control is the priority, and learning based control is a bonus, with a research team covering reinforcement learning. This is a hands on role with no direct reports to begin with, reporting to the VP Engineering, with a path to build a controls team as the company scales.
Key responsibilities
- Own locomotion and whole body control, and the body state estimation from IMU, joints and contact that it runs on.
- Own low level actuator control and system identification of actuators and contact.
- Set measurable targets for tracking, stability, disturbance rejection and latency, and prove them on hardware.
- Tune, debug and validate controllers on the robot, working with the Simulation and Systems Integration teams.
- Work with the research team on how learned policies and model based control fit together.
- Model predictive control or whole body optimization based control, deployed on physical hardware with measured results.
- Solid grounding in rigid body dynamics, forces, contact stability and latency.
- Strong C++ and Python.
- Experience with state estimation for legged or floating base robots.
- A graduate degree in controls or robotics, or around 8 to 10 or more years of hands on controls work on physical robots.
- Legged robot experience, strongly preferred.
- Reinforcement learning, as a bonus rather than a requirement.
- Trajectory optimization and reduced order models for legged locomotion.
- Torque controlled or quasi direct drive actuators.
- Footstep planning and integration with terrain perception.
- Tuning hardware and controllers for repeated impact loading.
- Tools such as Drake, Pinocchio or MuJoCo.
- A clean slate architecture, with the freedom to build controls properly from scratch.
- Early equity.
- The chance to own a domain and build a team around it.
