Case Study: GSK improves manufacturing processes and product quality by unlocking historical process data with MathWorks

GSK Uses Historical Data to Improve Manufacturing Processes and Make Better Products

GSK faced the challenge of extracting insight from a large archive of unsegmented time‑series data from toothpaste batch manufacturing (Sensodyne is regulated in many markets), with many isolated data sources and day‑to‑day variation. To tackle this, GSK worked with MathWorks — including MathWorks Consulting Services — using MATLAB and toolboxes such as Deep Learning Toolbox, Signal Processing Toolbox, and Statistics and Machine Learning Toolbox to turn unused process data into actionable information.

MathWorks developed a solution that automatically cleans, segments, and interrogates the time‑series data and wrapped it in a web app/GUI so nontechnical users can visualize and compare datasets. The MathWorks implementation linked disparate files and formats, enabled rapid code iteration with consulting support, and allowed GSK to reliably deduce mixer phasing and performance without human intervention, accelerating experiment design and process improvement.

GSK

Bob Sochon

GSK