Real-time quality control system for high-variability production at Estrella Galicia, built to verify bottle caps across formats, shapes, and colors on a continuous line. It includes:

• Vision-based inspection for real-time cap detection
• AI-driven model for robust recognition under high variability
• Integration with PLC systems for rejection and process control
• High-speed operation aligned with cycle times

The system combines traditional vision with machine learning to ensure consistent detection across heterogeneous conditions. Focus was on low latency, reliable decisions, and smooth integration with plant controls.

Challenges included high variability in bottle formats, colors, and geometries, strict cycle-time requirements, and integration with PLC infrastructure and rejection mechanisms.

The result is a reliable inspection system maintaining consistent quality and accuracy across product ranges, with stable operation at production speeds.