Robotics Engineer, Locomotion

Auxo Talent; San Francisco (CA); On-site; USD 150,000 - 230,000; Full time

Job summary

Auxo Talent in San Francisco on-site is seeking a Locomotion Engineer to build and deploy a locomotion stack on a physical humanoid robot. You will train policies, work on balance and whole-body control, and reduce the sim-to-real gap.

Ideal candidates have 1+ year applying locomotion policies to real robots, strong Python/C++, and familiarity with MuJoCo/Isaac Lab. Relocation to SF Bay Area is welcome; US citizenship or green card required.

Qualifications

• Minimum 1 year of professional experience deploying locomotion policies on physical legged robots.
• Strong Python and C++ programming skills.
• Working knowledge of MuJoCo, Isaac Lab or equivalent.
• An understanding of control theory beyond using a library.
• Based in the San Francisco Bay Area or willing to relocate; on site.
• Must be a US citizen or green card holder; visa sponsorship is not available.

Responsibilities

• Train locomotion policies with reinforcement learning and deploy on hardware.
• Work on whole body control, balance and recovery.
• Close the sim-to-real gap through system identification and domain randomisation.
• Combine learned policies with model-based control where appropriate.
• Report directly to the founders and collaborate with the team.

Skills

Python

C++

Robotics

Reinforcement learning

MuJoCo

Hardware deployment

Tools

MuJoCo

Isaac Lab

Job description

San Francisco | On site | $150,000 to $230,000 base

If you have trained a policy in simulation and then watched it fall over on hardware, you already understand this role.

We are working with a humanoid robotics company in San Francisco who need a Locomotion Engineer to build their locomotion stack and get it onto a physical humanoid. Simulation is a tool here, not the destination.

What you would be doing

• Training locomotion policies with reinforcement learning and getting them deployed on hardware
• Working on whole body control, balance and recovery
• Closing the sim to real gap through system identification, domain randomisation and whatever else the robot demands
• Combining learned policies with model based control where that is the right answer
• Reporting directly to the founders while collaborating with a larger team

What we need to see

• Minimum 1 year of professional experience deploying locomotion policies on physical legged robots. Simulation only experience will not clear the bar
• Strong Python and C++
• Working knowledge of MuJoCo, Isaac Lab or equivalent, alongside real hardware
• An understanding of control theory that goes beyond calling the library
• Based in the San Francisco Bay Area or willing to relocate. This role is on site
• Must be a US citizen or green card holder. This client is not able to sponsor or transfer visas

Why this one is different

There is a lot of impressive footage of robots dancing. Very little of it survives contact with an uneven floor. This team is building for the floor.