About Detect Defects
This site explores practical applications of computer vision and AI in quality control. From manufacturing defects to infrastructure inspection, we focus on real implementations, technical breakdowns, and honest assessments of what works (and what doesn’t).
What You’ll Find Here
Technical Deep Dives
How YOLO works for defect detection. Training models with limited data. Real-world performance benchmarks.
Industry Applications
Case studies from PCB inspection, textile manufacturing, structural monitoring, and assembly line automation.
Tool Reviews
Hands-on testing of Cognex, Landing AI, Roboflow, and other vision platforms. No affiliate bias—just practical insights.
Weekly AI Signals
Curated news on autonomous systems, enterprise AI, and emerging technologies that impact quality control.
About James Lions
I've spent the last several years working with computer vision systems in manufacturing environments. Most of that time has been debugging why models that worked perfectly in the lab failed on the factory floor.
This site started as notes to myself about what actually works in production. It turns out other people found them useful, so here we are.
I'm particularly interested in:
- Building defect detection systems that run reliably for months, not days
- Getting acceptable accuracy with limited training data
- Edge deployment challenges (spoiler: there are many)
- Where autonomous AI agents might actually be useful vs. pure hype
Not everything here will be cutting-edge. Some of the best solutions are boring and proven. That's fine by me.
Get In Touch
Have a question about a specific implementation? Spot an error in one of the tutorials? Drop me a line.
Building reliable vision systems, one defect at a time.