Sensor Network Robustness Using Model-Based Data Reconciliation for Continuous Tablet Manufacturing.

Area of Science:

• Pharmaceutical Manufacturing
• Chemical Engineering
• Process Control

Background:

• Continuous manufacturing in pharmaceuticals requires reliable process monitoring.
• Sensor technologies inherently have measurement errors and potential outliers.
• Robust monitoring is crucial for ensuring operational reliability in drug production.

Purpose of the Study:

• To demonstrate data reconciliation (DR) and gross error detection as real-time process management tools.
• To achieve robust process monitoring in pharmaceutical continuous manufacturing.
• To mitigate random measurement errors and detect nonrandom sensor malfunctions.

Main Methods:

• Developed and implemented model-based steady-state data reconciliation.
• Applied methods to two continuous tableting lines: direct compression and dry granulation.
• Utilized redundant sensor networks including embedded sensors and PAT tools.

Main Results:

• Successfully demonstrated DR and gross error detection on continuous tableting lines.
• Addressed challenges posed by process nonlinearity for real-time optimization.
• Validated the framework using at-line and off-line measurements.

Conclusions:

• Data reconciliation and gross error detection are effective real-time tools for pharmaceutical continuous manufacturing.
• These methods enhance process reliability by addressing sensor data quality.
• The demonstrated framework is applicable to diverse continuous manufacturing setups.