Autonomous fiber-optic cable insertion with a UR5e in simulation: board registration, learned SFP and SC port detection, multi-camera fusion and contact-based seating. My qualification-phase entry for Intrinsic’s AI for Industry Challenge, built on ROS 2 and PyTorch.
computer-vision robotics pytorch gazebo robotics-competition pose-estimation ros2 intrinsic robot-manipulation keypoint-detection force-control ur5e peg-in-hole cable-insertion ai-for-industry-challenge
-
Updated
Oct 18, 2026 - Python