Research Scientist, Safety-Critical Control, Robotics, SAF Lab

Amazon•Pasadena, CA

6d•$136,000 - $184,000•Onsite

About The Position

We are seeking a Research Scientist to join the SAF Lab. In this role, you will develop the core Control Barrier Function (CBF) theory and algorithms that form the mathematical foundation of the universal safety layer. Key to this process is a feedback loop between theory and practice: developing theory that is deployed on next generation robots and using experimental evaluation to drive new theory. This will enable you to push the boundaries of CBF theory: layered safety filters and trade-offs between robustness and optimality. A key challenge will be to understand the interplay with CBF theory and learned control policies, constructing safety filters that internalize learned policies and utilizing CBFs in learning to internalize safety. You will work with the inventor of control barrier functions and a team contributing directly to the next generation of CBF theory and its practical deployment across Amazon's diverse robot fleet.

Requirements

PhD in engineering, technology, computer science, machine learning, robotics, operations research, statistics, mathematics or equivalent quantitative field
Deep expertise in Control Barrier Functions, including theoretical foundations and practical implementation
Experience formulating and solving optimization-based controllers (QPs, SOCPs) for real-time safety filtering
Strong mathematical background in dynamical systems theory, nonlinear control, and formal verification or reachability analysis
Proficiency in C++ and Python with experience implementing control algorithms for real-time systems
Experience validating safety-critical algorithms on physical robotic hardware (not simulation-only)
Publication record at relevant venues (e.g., CDC, ACC, L-CSS, ICRA, RSS, RAL, Automatica, TAC, TRO)

Nice To Haves

Understanding of control systems engineering, with a specific focus on layered architecture used in robotic systems (high level planning, mid-level trajectory generation and low-level feedback control)
Familiarity with Hamilton-Jacobi reachability analysis and its relationship to CBF-based approaches
Experience with robust or adaptive CBF methods that account for parametric uncertainty or unmodeled dynamics
Experience with sum-of-squares (SOS) programming, Lyapunov function synthesis, or other computational tools for verifying set invariance
Familiarity with real-time embedded systems and the constraints of deploying optimization-based controllers on safety-rated hardware
Experience applying CBFs to multi-agent systems or high-dimensional robotic platforms (manipulators, legged robots)
Exposure to learning-based approaches for CBF synthesis (e.g., neural CBFs, data-driven barrier functions) and the integration of CBFs into learning (e.g., CBF-RL)

Responsibilities

Develop and implement novel CBF algorithms that provide formal safety guarantees while minimizing conservatism to maximize the permissible operating envelope highly dynamic robots
Frame safety filtering within complex layered architectures involving learning-based components, including VLAs, RL-based locomotion and whole-body controllers
Design multi-layer CBF based safety filters, including decision making layers, MPC, and real-time nonlinear feedback control elements
Formalize the interplay between models used in the CBF safety filter and the full order dynamics of the robotic systems, establishing formal guarantees even if the full order system dynamics is not known and contains learning-based elements
Understand the role of perception and semantic representations in the synthesis of CBFs, and the interplay between CBFs
Characterize the trade-offs between optimal safety and robustness to sensor noise, perception error, actuator and sensor failure
Address the theory-to-practice gap by developing CBF methods that are robust to model uncertainty, sensor noise, actuation delays, and computational latency
Implement real-time optimization solvers (e.g., QP-based safety filters) that execute within the tight timing budgets of safety-critical control loops
Validate algorithms through rigorous simulation and hardware experiments, characterizing failure modes and quantifying safety margins
Contribute to the theoretical foundations of CBFs through publications at top-tier controls and robotics venues
Collaborate with perception, planning, locomotion, and manipulation teams to ensure CBF formulations accommodate the needs of upstream and downstream systems
Collaborate with product teams and science leaders to set a science roadmap (with eventual impact on real robots)

Benefits

Medical, Dental, and Vision Coverage
Maternity and Parental Leave Options
Paid Time Off (PTO)
401(k) Plan
sign-on payments
restricted stock units (RSUs)
health insurance (medical, dental, vision, prescription, Basic Life & AD&D insurance and option for Supplemental life plans, EAP, Mental Health Support, Medical Advice Line, Flexible Spending Accounts, Adoption and Surrogacy Reimbursement coverage)
401(k) matching
paid time off
parental leave