Featured Projects

Machine Vision & ML Systems in Production

This page showcases selected projects from my work in machine vision, deep learning, and software engineering for manufacturing automation.

🤖 Vision-Guided Manufacturing Systems (2023-2025)

Slider Bar Transfer Equipment Vision System

Real-time measurement and guidance for slider bar assembly

Challenge: Develop a vision system to locate, measure, and guide the transfer of uncovered slider bars with high precision and reliability.

Solution:

  • Designed rule-based measurement algorithms for real-time geometric analysis
  • Implemented deep learning classifier for tray type identification
  • Built vision sequences synchronized with hardware-triggered cameras
  • Created auto-calibration routines to minimize setup time

Technologies: C#, Halcon, Deep Learning, Camera Triggering, Robot Integration

Impact: Enabled fully automated slider bar transfer

Hard Drive Assembly Vision System

High-speed vision for precision assembly guidance

Challenge: Provide real-time measurement and guidance for hard drive component assembly requiring sub-millimeter precision.

Solution:

  • Developed high-speed image processing pipeline for real-time robot guidance
  • Integrated vision system with robot control for closed-loop feedback
  • Designed robust auto-teach and calibration procedures
  • Implemented CI/CD pipeline for seamless software deployment

Technologies: C#, Machine Vision, Robot Control, GitHub Actions, CI/CD

Impact: Reduced assembly time by 30% while improving yield

Auto Glue Dispenser Vision System

Vision-guided precision glue dispensing

Challenge: Automate glue dispensing with vision feedback to ensure consistent quality.

Solution:

  • Conducted proof-of-concept using Halcon for vision-guided dispensing
  • Combined rule-based localization with ML-based anomaly detection
  • Automated robot axis movement based on glue droplet measurements
  • Implemented feedback loop for real-time quality control

Technologies: Halcon, Deep Learning, Vision-Guided Robotics, Anomaly Detection

Impact: Achieved consistent glue dispensing quality with minimal waste

🧠 Deep Learning in Production (2021-2022)

Wafer Inspection Deep Learning Module

Production-grade ML for semiconductor inspection

Challenge: Integrate deep learning into existing wafer inspection system for defect detection.

Solution:

  • Developed and trained defect detection models using Cognex VidiSuite
  • Designed robust deployment pipeline with extensive testing protocols
  • Created fallback mechanisms for edge cases
  • Provided production support and continuous improvement

Technologies: Cognex VidiSuite, Deep Learning, Image Processing, Production ML

Impact: Detected defects that were previously missed by rule-based methods

🔧 3D Vision Systems (2021)

Wafer 3D Inspection Module

Challenge: Develop comprehensive 3D inspection capabilities for wafer manufacturing.

Solution:

  • Integrated laser profiling and confocal microscopy for 3D capture
  • Built real-time 3D visualization UI for operator feedback
  • Optimized algorithms for high-throughput inspection
  • Conducted rigorous reliability testing across production scenarios

Technologies: C++, Laser Profiling, Confocal Microscopy, 3D Visualization, Image Processing

Impact: Enabled new quality control capabilities for complex 3D features

⚙️ CI/CD & DevOps (2023-2025)

Department-Wide CI/CD Implementation

Automated deployment for manufacturing software

Challenge: Modernize software deployment process across multiple production stations.

Solution:

  • Designed CI/CD pipelines using GitHub Actions for automated testing and deployment
  • Created common library packaging and versioning strategy
  • Implemented automated testing frameworks (unit tests, integration tests)
  • Advocated for CI/CD adoption across teams, conducting training sessions

Technologies: GitHub Actions, CMake, Python, Automated Testing, DevOps

Impact: Reduced deployment time from days to hours, improved software quality

🔬 Academic Research (2016-2020)

CUDA-Accelerated Optical Coherent Methods

GPU-accelerated phase retrieval for optical applications

Research Focus: Developed CUDA-based acceleration for phase retrieval techniques in optical measurement systems.

Contributions:

  • Implemented GPU-accelerated algorithms achieving 50x+ speedup
  • Published research on surface, displacement, and strain measurement techniques
  • Developed novel phase retrieval algorithms for challenging optical scenarios

Technologies: CUDA, C++, MATLAB, GPU Computing, Optical Engineering

Publications: Multiple peer-reviewed journal articles on optical metrology

🎓 Skills Demonstrated Across Projects

Machine Vision: Camera calibration, vision-guided robotics, real-time image processing, 3D reconstruction

Machine Learning: Deep learning deployment, anomaly detection, classification, model optimization

Software Engineering: Clean architecture, design patterns, SOLID principles, refactoring

DevOps: CI/CD pipelines, automated testing, version control, deployment automation

Programming: C++, Python, C#, CUDA, MATLAB

Tools: Halcon, Cognex VisionPro, OpenCV, TensorFlow, PyTorch, GitHub Actions

📫 Interested in Collaboration?

If you’re working on similar problems or want to discuss machine vision, ML in production, or software architecture for manufacturing, I’d love to connect!