PROCESS PERFORMANCE MONITORING – TOWARDS MODEL BASED APPROACHES
Systematic methods and tools for managing the complexity
Keynote Lectures: Theme-4
Keywords: MSPC; Performance Monitoring; Process Dynamics
Quality and consistency are key factors in determining business success. Manufacturing products that satisfy product quality and consistency specifications first time result in increased productivity and lower overall manufacturing costs. Approaches to achieving consistently high quality production and enhanced manufacturing performance include Statistical Process Control (SPC) and Lean and Six Sigma with increasing attention now being paid to Multivariate Statistical Process Control methodologies (MSPC) alongside the recent FDA Process Analytical Technologies (PAT) initiative focusing on Quality by Design.
In today’s process manufacturing environment, a number of issues arise which can challenge the application of MSPC based process performance monitoring. For example, most applications of MSPC have tended to focus upon the manufacture of a single product, i.e. one grade, one recipe; single site operations, etc. with separate models being developed to monitor individual product types or different production sites. In addition most methodologies are based on static PCA or PLS approaches with only a few addressing the impact of process dynamics and the impact of auto-correlated data. With process manufacturing trends being influenced by customer demands and the drive for product diversification, there is an increasing move towards responsive and flexible process manufacturing and hence dynamic performance monitoring.
The impact that process dynamics can have on the assured detection, diagnosis and sensitivity and robustness of the performance monitoring charts is becoming important. The introduction of dynamic model-based approaches such as Canonical Variate Analysis (CVA) and Autoregressive (AR) time series models into multivariate statistical process control based performance monitoring systems will be highlighted. These and other issues will be discussed and solutions proposed through demonstrations of industrial manufacturing applications in both batch and continuous processing. A number of industrial case studies will be presented.
Presented Tuesday 18, 17:05 to 17:45, in session Keynote Lectures: Theme-4 (T4-K1).
