Case Study: Auckland University of Technology improves milk powder quality and detects process faults with MathWorks' MATLAB machine learning

Auckland University of Technology and University of Auckland Researchers Analyze Dairy Processing Data with Machine Learning

Auckland University of Technology researchers at the Industrial Information and Control Centre (I^3C), working with industry partner Fonterra, tackled inconsistent functional properties of milk powder produced across multiple plants. Using MathWorks tools—primarily MATLAB along with the Statistics and Machine Learning Toolbox and Image Processing Toolbox—they needed to clean and align millions of inconsistent time‑series records from three plants over six years to determine when and why product quality varied.

Using MathWorks’ MATLAB, the team preprocessed and synchronized disparate datasets, interpolated missing values, ran PCA/PLS analyses, and used the Classification Learner to evaluate models; after resampling the training data they found boosted trees that delivered almost 95% prediction accuracy. MathWorks tools also enabled automated reporting, rapid evaluation of 20+ classifiers, image‑based particle analysis, and the identification and correction of a key process fault at a partner plant, yielding faster root‑cause insight and scalable handling of large datasets.

Auckland University of Technology

Nick Depree

Project Manager