Amazon
Sr. Applied & Research Scientist - Safe Autonomy Frontiers (SAF) Lab, Safe Autonomy Frontiers (SAF) Lab
$167K–$226K
Pasadena, CA · Senior · Full-time · Posted Oct 5, 2026
- Experience
- 3+ years for the Master's pathways; a PhD is an alternative.
- Education
- PhD, or Master's degree and 3+ years of relevant experience in computer science, robotics, control, mechanical engineering, electrical engineering, or a related field.
About the job
Develops and deploys safe-autonomy methods for dynamic robots, integrating control theory with learning and perception and validating research on physical hardware.
What you'll do
- Advance safe-autonomy research from formal foundations through integration with learning and perception to hardware validation.
- Test and validate next-generation robotic hardware, including locomotion, manipulation, and loco-manipulation.
- Publish research and collaborate with SAF Lab and Amazon production teams to deploy robotics research at scale.
What they ask for
- PhD, or Master's degree and 3+ years of industry or academic research experience; the posting also specifies a Master's degree and 3+ years of applied science, research, or robotics engineering experience, or a PhD, in a relevant field.
- Hands-on experience developing and deploying control algorithms and/or learning policies on physical robotic hardware.
- Working knowledge of safety-critical control, including control barrier functions and safety filters.
- Proficiency in C++ and Python, plus experience with robotics physics simulators and demonstrated technical impact through shipped systems, patents, or top-tier publications.
Nice to have
- Professional software development experience and experience taking robotics research from prototype to deployed product.
- Knowledge of learning-based robotics, perception, control architectures, reachability, and/or real-time deployment constraints.
Pay
167,100.00 - 226,100.00 USD annually (base salary)
Tools
C++, Python, Isaac Gym/Sim, MuJoCo, PyBullet, MPC, MPPI, LiDAR
AI focus
Safe reinforcement learning for agile whole-body control and layered safety filters integrated with learning, perception, and semantic reasoning.
Model-written summary · the original posting is authoritative