Lead Scientist, Contact-Rich Robotic Manipulation

Amazon; North Reading (MA); On-site; USD 199,000 - 269,000; Full time

Benefits offered by this job

Health insurance

RSUs

401(k) matching

Job summary

Amazon's Manipulation Robotics team seeks a Principal Applied Scientist to define control architectures for next-generation grasping systems in cluttered environments. You will own the technical direction for force- and impedance-based manipulation across production-scale tasks, collaborating with scientists, hardware, and systems engineers to deliver scalable, robust solutions.

You will mentor engineers, publish and review technical work, and ensure a cohesive platform design that couples

Qualifications

• PhD or equivalent body of published work in robotics, ML, CS, EE, ME, or related field.
• Experience writing production-quality Python and C++ code for real robotic hardware.
• Hands-on expertise in force, impedance or admittance control for contact-rich manipulation.
• Experience integrating learned policies with low-level controllers for contact-rich tasks.

Responsibilities

• Own the technical direction and contact-control strategy for contact-rich manipulation, including the boundary between low-level control and learned policy, balancing tradeoffs among speed, performance, quality, cost, complexity, and adaptability.
• Identify open scientific problems that gate contact-rich manipulation at production scale, and invent methods to solve them.
• Architect contact control built on force and tactile sensing, covering compliant-contact behaviors, hybrid position and force control, and the safe interaction envelope; write critical-path code.
• Partner with scientists, hardware designers, and systems engineers to treat platform, arm, and end-effector as a single hierarchical system, co-designing compliant end-effectors.
• Create mechanisms to learn from fielded production and reason about failure modes to improve recovery from unexpected contact.
• Drive how simulation, analytical models, demonstrations, and real-robot data are used to scale warehouse performance from lab results.
• Establish reference implementations, evaluation standards, and design-review practices enabling others to extend the architecture.
• Develop other scientists and engineers through mentorship, technical review, and leadership in hiring.

Skills

Python

C++

Force control

Impedance control

Technical leadership

Education

PhD in Robotics or related field

Job description

Amazon's Manipulation Robotics team seeks a Principal Applied Scientist to define control architectures for next-generation grasping systems in cluttered environments. You will own the technical direction for force- and impedance-based manipulation across production-scale tasks, collaborating with scientists, hardware, and systems engineers to deliver scalable, robust solutions.

You will mentor engineers, publish and review technical work, and ensure a cohesive platform design that couples