Industrial Automation
Industrial Automation Projects
Production lines run product that looks identical to the eye but reads as a different geometry to a computer. These are the vision, control, and commissioning systems I've built to measure and act on that variation at line speed.
PROJECT 01
Inline Bun Quality Inspection System
Developed a full-stack machine vision inspection system for inline bun quality control, using a Time-of-Flight 3D camera to compare each product's point cloud against a reference model in real time.
<500 ms
Scan-to-decision time
3 months
Archived inspection history
My Role
End-to-end development — custom Windows application for data collection (C#), 3D image processing (Halcon), and PLC integration for pass/fail sorting.
Technologies
C#, Halcon, Time-of-Flight camera, Allen-Bradley PLC, TCP/IP.
Focus
Sub-500 ms scan-to-decision time, point-cloud scoring, and three-month result and image archiving for traceability.
PROJECT 02
Fries Length & Quality Grading System
Designed the operator and manager HMI for a production-line fries inspection system, integrating live camera imaging to monitor length ratio and defect rate on an hourly basis.
120 pc/min
Inspection throughput
±1%
Measurement accuracy
Hourly
Length & defect quality checks
My Role
HMI design and development (Ignition), Gocator linescan camera integration, and conveyor/motor speed configuration to optimize image quality.
Technologies
Ignition, Gocator linescan camera, PLC.
Focus
Simple on/off controls for operators, configurability and history tracing for managers, and consistent product quality across production runs.
PROJECT 03
Freezer System Commissioning
Commissioned a production freezer system end-to-end, from control panel I/O verification through motor and belt tuning, VFD configuration, and full network integration with remote panel access.
My Role
Controls commissioning — I/O checkout, PID and motor ratio tuning, VFD configuration, and network/remote access setup.
Technologies
PLC, VFD, PID control, Industrial Ethernet, VPN/remote access.
Focus
Control panel I/O verification, motor and belt speed tuning, VFD parameter validation, and secure remote access to the plant network.
PROJECT 04 · GRADUATE TEAM PROJECT
Dew Point Prediction for Compressed Air Systems
Predicting dew point in a manufacturing plant's compressed air drying system, to move from reactive threshold-based control toward proactive, predictive control.
0.9925
R² on test set
0.0089
RMSE on test set
My Role
Implemented and compared regression algorithms in Python, and led hyperparameter tuning on the best-performing model.
Technologies
Python, pandas, scikit-learn, GridSearchCV.
Focus
Feature engineering, comparative evaluation across eight regression models, and hyperparameter tuning validated by R² and RMSE.
Correlation between dew point and the four most predictive features — supply temp, discharge temp, step, and motor amperage.