Vision-Based Robotic Drilling for Aerospace Manufacturing

Research on precision robotic drilling with online error correction using machine vision.

Stage
Research Prototype
Published
Who made it?
Ali Maghami
Tags
Computer VisionRoboticsDeep LearningAerospaceResearch
Share

Overview

Conducted research on vision-based guidance systems for precision robotic drilling in aerospace composite manufacturing, translating computer vision research into practical systems for industrial applications.

Contributions

  • Engineered custom optical measurement system for online robot path correction
  • Developed target-localization methods enabling real-time process adaptation
  • Created automated multi-light imaging workflows with deep learning for defect detection in composites
  • Published peer-reviewed research demonstrating practical applicability to manufacturing

A crosshair laser profiler recovers the surface geometry ahead of the tool. Laser lines are tracked inside a moving window on each frame, transformed from the camera frame into the laser-plane frame, and interpolated into a 3D spline that the robot path is corrected against online.

Technologies

OpenCV, PyTorch, Semantic Segmentation, ROS, KUKA Robot Control, Laser Profiling