Senior Robotics Controls Engineer

Confidential; San Francisco (CA); On-site; USD 150,000 - 220,000; Full time

Job summary

Confidential in San Francisco Bay Area is seeking a Senior Robotics Controls Engineer to design and tune core control systems for 200+ ton autonomous haul trucks operating in remote mining sites. You will own localization, path following, longitudinal and lateral control, and steering.

This is a production role, not a research one. You will work with perception, planning, and hardware teams to ship into production vehicles, test in the field, and ensure robustness when GPS drops out.

Qualifications

• 3–10 years of robotics controls engineering experience.
• Localization, lateral/longitudinal control, or navigation deployed on real autonomous vehicles.
• Experience designing and tuning MPC controllers for production systems.
• Strong Python and C++ for production control systems.

Responsibilities

• Design, implement, and tune MPC and PID controllers across varying payloads and terrain.
• Build localization pipelines fusing GPS/RTK, IMU, and other sensors.
• Debug control issues through field testing and log replay.
• Collaborate with perception, planning, and hardware teams to ship into production vehicles.

Skills

ROS

MPC tuning

Python

C++

Localization

Autonomous vehicles

Production deployment

Education

BS/MS/PhD in Robotics or related engineering

Tools

NVIDIA Jetson

CAN Bus

NumPy

SciPy

Job description

Sponsorship: open to visa transfers (OPT, H1B) and new sponsorships (H1B, TN)

We're hiring a Senior Robotics Controls Engineer at a well-funded Physical AI company running one of the largest autonomous mining fleets in the world. You'll design and tune the core control systems that let 200+ ton autonomous haul trucks operate safely in remote, unstructured mining sites, owning localization, path following, longitudinal and lateral control, and steering.

This is a production role, not a research one. The company has 100+ autonomous trucks in live operation today, hauling millions of tons for real customers. Your code runs on real vehicles, gets tested in the field, and has to work when GPS drops out in the middle of a remote mine.

The problem you'd help solve:

Mining sites are some of the hardest environments for autonomy: no road markings, shifting terrain, huge variation in payload and grade, and unreliable GPS. Keeping a 200+ ton truck on its path in those conditions takes robust control and state estimation that holds up outside the lab. This team has already proven it works at scale, and now needs strong controls engineers to push performance further and scale the fleet.

You'll design, implement, and tune MPC and PID controllers across varying payloads and terrain, build localization pipelines fusing GPS/RTK, IMU, and other sensors, debug control issues through field testing and log replay, and work with perception, planning, and hardware teams to ship into production vehicles. Expect periodic travel (up to 15%) to customer mining sites for deployment and tuning.

You'll likely be a fit if you have:

• Hands-on experience with ROS and Linux (required)
• Hands-on experience designing and tuning MPC controllers
• 3-10 years of experience in robotics controls engineering
• Localization, lateral/longitudinal control, or navigation deployed on real autonomous vehicles
• Strong Python and C++ for production control systems
• A BS, MS, or PhD in Robotics, Mechanical, Aerospace, or Electrical Engineering
• A fast, high-urgency execution style: you ship to production, not just papers

Nice to have:

• Autonomous mining experience (e.g. Caterpillar, Komatsu, Hexagon, Freeport)
• Autonomous vehicle experience at companies like Aurora, Tesla, Uber, or Nuro
• An MS or PhD in a controls-related discipline
• Experience with NVIDIA Jetson, CAN Bus, NumPy, SciPy

What you won't find here:

This role won't suit you if your background is purely academic or research with no industry deployment, if you want remote or hybrid work, or if you'd rather work in simulation than on real hardware. You'll be on-site 5 days a week (with flexible hours), in the field on real trucks, and expected to own control problems end-to-end rather than hand them off.