AI-Powered Automated Visual Inspection of IV Bags: A Real-World Case Study: Poster Presented at PDA Week 2026

Area of Science:

• Medical Device Manufacturing
• Artificial Intelligence in Quality Control
• Automated Visual Inspection

Background:

• High-volume IV bag lines experienced frequent false rejects due to printing variability.
• Existing inspection methods struggled with print artifacts, bag deformation, and lighting inconsistencies.

Purpose of the Study:

• To develop and implement an AI-powered automated visual inspection (AVI) system for IV bag production.
• To reduce false rejects and improve the accuracy of particulate detection.
• To maintain print legibility and enhance line efficiency.

Main Methods:

• The AVI system utilizes deep learning for defect detection and high-quality image capture.
• Integration with existing Manufacturing Execution Systems (MES) and packaging controls.
• Addressing practical design considerations: camera placement, illumination, data management, and model training.
• Ensuring regulatory alignment through Installation Qualification/Operational Qualification/Performance Qualification (IQ/OQ/PQ), change control, and Good Manufacturing Practices (GMP).

Main Results:

• The AVI system successfully learned to differentiate subtle print artifacts from true particulates in real time.
• Reduced misclassifications caused by printing variability, bag deformation, and lighting.
• Demonstrated seamless integration without slowing production throughput.

Conclusions:

• AI-enabled AVI systems can effectively address challenges in IV bag inspection, reducing false rejects.
• The developed system enhances detection of particulates on flexible bag formats and improves overall line efficiency.
• Lifecycle management, including retraining and performance monitoring, is crucial for sustained AI inspection performance in continuous production.