Webinar on
A framework for the deployment of soft-sensors and multi-variate process
monitoring in regulated process industries
John Mack and Ewan Mercer, Applied Materials
3rd July 2025 at 3pm (UK time)
A framework for the
deployment of soft-sensors and multi-variate process monitoring in regulated
process industries
Soft-sensors are
mathematical models used to predict critical process parameters (CPPs) or
critical quality attributes (CQAs) in industrial processes. Typically, a soft
sensor is used where the CPP/CQA cannot be measured in line/at-line and is a
critical part of a process control strategy. Multivariate statistical models
are commonly used to monitor a process, a specific piece of equipment or unit
operation.
In the case of soft
sensors, the mathematical model may be a machine learning model, a mechanistic
model or a hybrid combination. For process monitoring, the multivariate
statistical model is nearly always a machine learning model. Regardless of the
model application, model type and structure, consideration must be given to
validation and scope of the model, the uncertainty in the prediction and
monitoring of the model performance.
In this webinar the
focus is on the deployment of these models within the regulated process
industries. The ISPE GAMP5 2nd Edition and ICH guidelines in
particular embrace a risk-based approach to model deployment. The webinar
explores model development, scope, uncertainty, deployment and performance
monitoring.
These considerations
are particularly important when the soft-sensor prediction is included in an
Advanced Process Control strategy.
This webinar will last
no longer than one hour.
Please register direct at:
https://universityofstrathclyde.webex.com/weblink/register/r7a04d056c14026ca910126debcf4d918
