Staff ML Scientist, Autonomous Systems Simulation

The challenge

A meaningful fraction of scenarios that pass in simulation fail on the real vehicle due to the sim-to-real gap. Closing this gap through better physics models, sensor simulation, and domain randomization techniques is the core research problem.

Role overview

You research and develop the simulation technologies that validate autonomous driving. Sensor simulation (LiDAR, camera, radar), physics modeling (vehicle dynamics, tire-road interaction), and scenario generation at scale. The research here directly determines how much of the safety validation can happen virtually versus on public roads.

Specializations

Technical requirements

• PhD in Computer Science, Robotics, Physics, or related field
• 10+ years in simulation, rendering, or computational physics, with 4+ years in autonomous systems
• Published research in simulation for robotics or autonomous vehicles
• Strong background in physics-based rendering and sensor modeling
• Proficiency in C++ and Python, with experience in GPU-accelerated computing (CUDA)
• Experience with simulation platforms (CARLA, LGSVL, internal tools)

Key responsibilities

• Research and develop physics-based simulation models for autonomous vehicle validation
• Improve sensor simulation fidelity for LiDAR, camera, and radar across weather and lighting conditions
• Design domain randomization strategies that improve sim-to-real transfer for perception and planning
• Develop automated scenario generation pipelines that produce safety-critical edge cases at scale
• Collaborate with the perception and planning teams to define simulation fidelity requirements
• Publish research and present at top conferences (RSS, CoRL, ICRA)

Compensation & Benefits

• RSU package with annual refresh
• Base salary: $250,000 - $300,000
• Research publication bonus ($2,000 per accepted paper)
• Sabbatical program (6 weeks every 4 years)
• Relocation assistance