Applied Scientist, Tactile Manipulation for Robotics

Amazon Inc.; Seattle (WA); On-site; USD 172,000 - 222,000; Full time

Benefits offered by this job

Health insurance

401(k) matching

Paid time off

Parental leave

Job summary

Amazon is hiring an Applied Scientist to develop tactile- and force-driven manipulation policies for dexterous robotic hands and grippers. You will research policy learning methods that enable grasping and in-hand manipulation in unstructured environments, focusing on sim-to-real transfer and real hardware deployment.

You will work across tactile sensing, manipulation, and multi-modal sensing, and mentor engineers while publishing in top venues.

Qualifications

• Experience programming in Java, C++, Python or related language.
• Research or applied experience in tactile sensing, manipulation policy learning, contact-rich manipulation, dexterous grasping, or multi-modal perception for robotics.
• Experience deploying learned policies or control algorithms on physical robotic hardware.

Responsibilities

• Design manipulation policies using tactile/force feedback for dexterous grasping and in-hand manipulation.
• Drive sim-to-real transfer between simulation and real hardware.
• Shape objectives considering actuator behavior, transmission ratios and thermal limits.
• Build simulation-based and on-robot evaluation frameworks with benchmarks for tactile sensing.
• Own projects within tactile sensing and manipulation workstreams; senior staff may set direction across a workstream.
• Collaborate across hardware, design, firmware, controls teams to deploy policies.
• Publish at top venues and collaborate with the external research community.
• Contribute to scientific standards via code and design reviews and mentoring.

Skills

Java

C++

Python

Tactile sensing

Policy learning

Tools

MuJoCo

Isaac

URDF

Job description

Amazon is hiring an Applied Scientist to develop tactile- and force-driven manipulation policies for dexterous robotic hands and grippers. You will research policy learning methods that enable grasping and in-hand manipulation in unstructured environments, focusing on sim-to-real transfer and real hardware deployment.

You will work across tactile sensing, manipulation, and multi-modal sensing, and mentor engineers while publishing in top venues.