Upstream & Downstream Analysis
Customer Challenges
Company wanted to use advanced analytics techniques to:
Complete multivariate analysis on Upstream Bio Reactor and downstream chromatography data
Complete Coefficient variability analysis to find key contributing model inputs
Rank key process variables
Modelling of best fit / golden batches to determine the scale of process variability
Provide recommendations for reduced variability of Titre & protein extraction levels
Solution
Phase 1 – Downstream analysis covering:
15 days professional service engagement over 3 months
Extraction of Chromatography and Cross Filtration data from Cytiva AKTA and UNIFLUX systems via UNICORN
Analysis of data to assess stability, controllability and predictability of downstream processing using TIBCO Spotfire and Sartorius SIMCA
Phase 2 – Upstream analysis covering:
30 days professional service engagement over 3 months
Extraction of SCADA data from CYTIVA Xcellerex / AVEVA (Wonderware), combined with Offline and Harvest data
Custom pre-processing code written to process 3,000+ extracted data files for 20 batches & prep for multi-variate analysis
Analysis of data to assess stability, controllability and predictability of Upstream processing using MATLAB and Sartorius SIMCA
Benefits
Identification of factors effecting quality & hypothesis for further manufacturing optimisation.
Related Case Studies
Food Industry
The client needed to augment their business intelligence team as they looked to update their existing Power BI / Tabular business intelligence architecture to a Power BI / Fabric environment.
Healthcare
Clarke Analytics completed an AS-IS and TO-BE analysis across a few key customer lifecycle processes to identify possible areas for improvement.
Pharma
8-week professional service engagement,
Data Analytics Maturity Assessment, interview with all key business units & IT data governors.
Let's explore the art of the possible