Postdoctoral Scholar - SAF Lab, Compass
Core
Research safe autonomy for highly dynamic robots (loco-manipulation, mobile manipulators) by integrating control barrier functions (CBFs) with perception and learning, validated on hardware.
Role type
Postdoctoral Scholar (Research Scientist)
Builds
Simulation/evaluation pipelines, sim-to-real transfer methods, and hardware-deployed safety controllers for robots operating alongside humans.
Domain
Robotics, Control Theory, Machine Learning
Deliverable
production ML models | research
Required skills
Control barrier functions (CBFs), safe reinforcement learning, C++ and Python, physics simulators (Isaac Gym/Sim, MuJoCo, PyBullet), hardware validation, top-tier publication record
Preferred skills
Locomotion, reduced order models, layered control architectures, nonlinear control, reachability methods, whole-body control, learning-based CBF synthesis, perception (LiDAR, depth cameras), Hamilton-Jacobi reachability, safety-constrained RL, model-based control (MPC, QP), hierarchical RL, real-time deployment constraints, large-scale RL infrastructure
Technologies
C++, Python, Isaac Gym/Sim, MuJoCo, PyBullet, CBFs, Reinforcement Learning, MPC, QP
Responsibilities
Push fundamental science of safe autonomy (theory, learning integration, perception synthesis); develop simulation and evaluation pipelines; build sim-to-real transfer pipelines; deploy methods on hardware and validate science; publish at top-tier venues; collaborate on science roadmap
Seniority
Postdoctoral Scholar (Early-career Researcher)
