System Integration · Google (via AutoRoboto)
Automated biometric testing for Pixel.
Robotics lab automation for evaluation of Google Pixel biometric face authentication models. I built the trajectory system and led the ROS1→ROS2 migration, deployed across 10 internal lab installations.
Note: This project page is intentionally high-level. For details
about the work, feel free to reach out directly.
What I work on
- Trajectory system: Built a trajectory system that varied robot pose, lighting and background conditions to stress-test biometric authentication models under controlled conditions.
- Capture-frame precision: Mapped robot position and velocity to the exact camera capture frame, enabling more precise and reproducible evaluations.
- Human motion on the robot: Measured human movements using an IMU and integrated those motions on the robot, broadening the range of authentication modalities the system could evaluate.
- ROS1 → ROS2 migration: Led the system-wide architecture migration: a threading architecture rewrite and improved motion planning, deployed across 10 internal lab installations.