Data-driven engineering framework with AI algorithm of Ginkgo Folium tablets manufacturing.

Area of Science:

• Pharmaceutical Manufacturing
• Data-Driven Engineering
• Smart Manufacturing

Background:

• Smart manufacturing presents significant challenges in pharmaceutical production.
• Existing methods often struggle to effectively diagnose and improve complex manufacturing processes.

Purpose of the Study:

• To propose an original data-driven engineering framework to address challenges in pharmaceutical smart manufacturing.
• To diagnose process capability and enhance quality traceability in Ginkgo Folium tablet production.

Main Methods:

• Characterization of nearly 2,000,000 real-world data points from Ginkgo Folium tablet manufacturing.
• Development of a digital process capability diagnosis strategy using multivariate Cpk integrated with Bootstrap-t.
• Analysis of quality traceability from unit to end-to-end levels.

Main Results:

• Identified weak process capability (Cpk) in Ginkgo Folium extracts (0.59), granules (0.42), and tablets (0.78), with granulating being particularly problematic.
• Quality traceability analysis showed improvement, decreasing from 2.17 to 1.73.
• The findings underscore the need to focus on the granulating stage to enhance overall product quality.

Conclusions:

• The proposed data-driven engineering strategy empowers industrial innovation for smart pharmaceutical manufacturing.
• Targeted improvements in the granulating process are crucial for enhancing the quality characteristics of Ginkgo Folium tablets.
• This framework offers a robust approach to diagnosing and improving process capabilities in pharmaceutical production.