Continuous process verification and Product Control Strategy

The Product Control Module creates context from heterogenous discrete and analog data sources to detect hidden CPV variability signals with machine learning to create a real-time dynamic product control view for the enterprise.

Build a real-time CPV program

A real-time CPV program can only be built on a foundation of CPV signals designed to identify potential new variation or unexpected patterns in the data while the process is operating within the operation space of the product control strategy. The existing process control systems and SPC are neither intended to, nor capable of,  measuring these signals.  A big data and machine learning solution addresses all these challenges.

  • Increases yield and profit from dynamic management of variation within specified limits defined in the control strategy. 
  • Provides evidence based demonstration of control strategy robustness in APR’s   . 
  • Reduces OOS and CAPA’s with early detection from CPV signals with a focus on a risk-based review by exception 
  • Reduces time, effort, and cost of deviation investigations  

Monitor the lifecycle of product performance with big data and machine learning

A typical product control strategy produces thousands of data points that span the lifecycle of the product and the manufacturing infrastructure of the enterprise including CMO’s and CDMO’s. The CPV signal discovery therefore requires understanding complex multi-dimensional, often non-linear, relationships and slow emerging patterns from a large volume of heterogenous data.

  • diversification

    Early indication of increasing risk 

    Identification and summarization of CQAs —and influencing CPPs and CMAs— from each unit operations contributing to risk 

  • graph

    Early indication of variability 

    Identification and summarization of causes of variability. Measurement and comparison of variability at product, Unit Procedure and phase levels 

  • analytics

    Product and enterprise-wide measurement  

    Identification and summarization of causes of variability. Variability can be measured across batches, campaigns and lots

  • analytics-1

    Detailed root cause identification with  MVDA  

    Identification and summarization of causes of variability. Detailed root cause identification with  MVDA

In general, CPV signals assess predicted performance based on previous process experience. The development and effectiveness of these signals depend on statistical techniques sensitive to the size and inherent variability of the existing data set.

— Biophorum International
The Quartic CPV and Product Control Strategy Solution is built on the Quartic Platform. The Platform connects to industrial OT (Operational Technology), condition measurement systems, MES and CMMS systems; contextualizes the data to an asset and allows users to build intelligence agents using machine learning (ML) and complex event (rules) processing (CEP).