Robotics Simulation Scientist — GPU RL & Sim-to-Real

Amazon; Bethesda (MD); On-site; USD 143,000 - 193,000; Full time

Job summary

Amazon’s Robotics Simulation team seeks an Applied Scientist to design, build, and validate simulation environments and policy training pipelines for manipulation and mobility. You will drive end-to-end simulation science deliverables and collaborate with SDEs, Technical Artists, and partner science teams.

The role focuses on GPU-accelerated RL environments, domain randomization, and sim-to-real transfer, with increasing independence and impact over time.

Qualifications

• PhD or Master’s degree in a relevant field.
• Experience in robotics design, automation systems, control systems, or related product development.
• 2+ years of experience with physics simulation platforms for robot learning (MuJoCo, Isaac Sim/Lab, PyBullet, Drake).
• Demonstrated experience training robot policies via reinforcement or imitation learning and evaluating in simulation.
• Familiarity with URDF/MJCF/USD formats and rigid/soft body dynamics.
• Knowledge of sim-to-real transfer concepts (domain randomization, system identification, physics calibration).

Responsibilities

• Design and implement GPU-accelerated RL and imitation learning environments for manipulation and mobility tasks.
• Build and maintain policy training pipelines for diverse model architectures and evaluate trained policies in simulation.
• Characterize and reduce sim-to-real gaps by comparing simulated vs real robot data and implementing improvements.
• Implement domain randomization strategies to improve policy robustness and transfer to real hardware.
• Develop sim-to-real transfer techniques including system identification and physics parameter calibration.
• Create embodiment validation tests to ensure digital twins match real hardware.
• Build data pipelines for recording, replaying, and augmenting demonstration data to scale training data.
• Contribute to end-effector modeling and contact dynamics tuning for realistic interactions.
• Author design documents and participate in technical reviews.
• Collaborate with partner science teams to meet training requirements.

Skills

Reinforcement learning

Imitation learning

Robotics simulation

Policy evaluation

Environment design

GPU computing

Education

PhD or Master’s degree

Tools

JAX

PyTorch

vLLM

SGLang

Dynamo

TorchXLA

TensorRT

Job description

Amazon’s Robotics Simulation team seeks an Applied Scientist to design, build, and validate simulation environments and policy training pipelines for manipulation and mobility. You will drive end-to-end simulation science deliverables and collaborate with SDEs, Technical Artists, and partner science teams.

The role focuses on GPU-accelerated RL environments, domain randomization, and sim-to-real transfer, with increasing independence and impact over time.