Applied Scientist, Amazon Robotics - Vulcan

Amazon Amazon · Big Tech · Seattle, WA · Applied Science

This role focuses on designing, implementing, and deploying motion planning, control, and manipulation algorithms for robotic systems in Amazon fulfillment centers. It involves both engineered and learned approaches, with a strong emphasis on sim-to-real deployment and validation on production systems at scale. The role contributes to learned manipulation behaviors and controllers, requiring production-quality code and scalable, real-time implementations.

What you'd actually do

  1. Design, implement, and deploy motion planning, control, and manipulation algorithms on production robotic systems.
  2. Partner with experts across disciplines including perception, hardware and software to create intelligent, integrated systems and solutions.
  3. Contribute to the development of learned manipulation behaviors and controllers, including sim-to-real deployment.
  4. Write production-quality code and own scalable, real-time implementations.
  5. Validate algorithms on hardware, iterating between simulation and real-world testing to ensure robust performance.

Skills

Required

  • PhD, or Master's degree and 4+ years of science, technology, engineering or related field experience
  • Experience programming in Java, C++, Python or related language
  • Experience with robotics work cells and control systems
  • Strong background in motion planning, control theory (compliant control, trajectory optimization), or learning-based manipulation

Nice to have

  • Experience deploying and supporting complex robotic systems at scale
  • Familiarity with reinforcement learning, behavior cloning, and/or sim-to-real transfer for manipulation
  • Experience with contact-rich manipulation, force/torque control, and/or constrained motion planning
  • Publications in top robotics, controls, or machine learning venues (RSS, CoRL, ICRA, IROS, NeurIPS, ICML, etc.)

What the JD emphasized

  • production robotic systems
  • learned manipulation behaviors
  • sim-to-real deployment
  • production-quality code
  • real-time implementations
  • contact-rich manipulation
  • force/torque control
  • constrained motion planning

Other signals

  • robotic systems
  • contact-rich manipulation
  • learned manipulation behaviors
  • sim-to-real deployment
  • production robotic systems