CPACT Webinar
on
From Data to
Control: AI-Enabled PAT and APC for
Fast
Development and GxP Manufacturing
Qiaolin
Yuan, Modersys
20th
August 2026 4pm Singapore time, 9am UK time
Interest in the application of Artificial Intelligence (AI) and Machine
Learning (ML) in manufacturing is expanding, particularly in pharmaceutical and
other regulated industries where Process Analytical Technology (PAT) and
Advanced Process Control (APC) are well established. However, adoption remains
limited by fragmented data environments, high experimental effort, and the challenges
associated with deploying data-driven models within
validated GxP systems.
This work presents a practical framework for
integrating AI with PAT and APC across the process lifecycle. The approach
combines a unified data infrastructure, linking equipment, PAT, and laboratory
systems, with hybrid modelling techniques that integrate mechanistic knowledge
and data-driven models. This enables more efficient experimentation, improved
process understanding, and the development of control strategies that can
transition from development to manufacturing.
Industry case studies from small and large molecule pharmaceutical
processes are used to illustrate implementation. These include adaptive APC
methods for rapid process characterisation prior to deployment of fixed models
in production, AI-assisted equipment setup, and multivariate APC applications
demonstrating reductions in process variability and improved stability of
critical quality attributes. While pharmaceutical case studies will be used to
illustrate implementation, the underlying concepts are transferable to other high-value
manufacturing sectors, including nutritional powders, food, bio-solutions, and
other process industries.
Governance of AI within GxP is a
central consideration. Model lifecycle management aligns with emerging
regulatory expectations, including:
· Risk-based classification by context of use;
· Structured validation;
· Continuous performance verification.
Transparency and traceability are supported through model documentation
and explainability, with human oversight retained for critical
decisions.
The framework also addresses data security and integrity through
controlled integration of equipment interfaces, including cloud-enabled
architectures, with auditability, access control, and version management
aligned to GMP requirements.
Overall, this work demonstrates a controlled approach to embedding AI
within PAT and APC workflows, supporting improved process robustness,
efficiency, and scalability.
This webinar will take no longer than one
hour.
The webinar is for CPACT members only and is
free to attend.
Please register at https://universityofstrathclyde.webex.com/weblink/register/rd12dfa0ba4343789e2b0d6a41a778f00