Applied Scientist II

Amazon Amazon · Big Tech · Sunnyvale, CA · Applied Science

The role focuses on developing and applying cutting-edge simulation methodologies for advanced robotics systems, including physics-based simulation, sim-to-real transfer, and machine learning. The goal is to enable rapid development, testing, and validation of robotic systems in complex environments. The role involves fundamental research and real-world development, translating research into scalable simulation capabilities that impact robot design and building.

What you'd actually do

  1. Advance physics-based simulation fidelity for contact-rich manipulation and locomotion
  2. Design and build high-performance simulation tools integrated into a robotics design stack
  3. Translate research ideas into robust, verifiable data
  4. Develop methods to quantify and reduce simulation-to-reality gaps across design, safety, and control
  5. Architect scalable simulation solutions for rigid and deformable body dynamics

Skills

Required

  • PhD in computer science, computer engineering, or related field
  • 2+ years of science, technology, engineering or related field experience
  • Deep expertise in physics-based simulation, including rigid and deformable dynamics, contact mechanics, computational geometry, and numerical methods
  • Experience designing and optimizing physics-based simulation systems for high-performance and large-scale computing environments
  • Strong programming skills in C++ and Python, with an emphasis on maintainable, performance-critical code
  • Working knowledge of modern physics engines such as MuJoCo, Isaac Lab, Drake, and Newton

Nice to have

  • Experience with reinforcement learning and policy training in simulation
  • Familiarity with differentiable physics, learned simulation models, or neural physics engines
  • Background in contact-rich manipulation or legged locomotion simulation
  • Experience with robotics model formats and pipelines (e.g., URDF, SDF, USD)
  • Expertise in GPU-accelerated computing and algorithms
  • Experience deploying simulation-trained policies on real robotic systems
  • Demonstrated research leadership, from project conception through publication and deployment

What the JD emphasized

  • Deep expertise in physics-based simulation
  • Experience designing and optimizing physics-based simulation systems
  • Experience with reinforcement learning and policy training in simulation
  • Background in contact-rich manipulation or legged locomotion simulation
  • Demonstrated research leadership

Other signals

  • Develop methods to quantify and reduce simulation-to-reality gaps
  • Translate research ideas into robust, verifiable data
  • Advance physics-based simulation fidelity