Contractual agreements between RAAL and their customers specify final product specifications, including the strength of crystals which is a key determinant of quality. RAAL use the Bayer process to extract alumina from a reddish brown ore called bauxite. This complex manufacturing process involves four different stages, and typically takes around 5 days. RAAL therefore wanted to improve the forecasting ability of their current models to predict alumina quality 5 days in advance.

Soft sensors use mathematical or statistical models to enrich the information measured by actual online sensors and offline sources, such as laboratory data. A team led by CHIMiRA mathematician, William Lee, trialled several models to drive a soft sensor for measuring product quality, including differential equation modelling, principal component analysis, stepwise regression, and time series analysis. Comparing the different methods against RAAL’s extensive historical process and laboratory data, the stepwise regression approach was shown to be the most effective.

The stepwise regression model of RAAL’s manufacturing process increased prediction accuracy of output quality by 200%, and also distinguished between normal operation and process events - such as power cuts - which significantly impact product quality. This model was integrated into the company's daily workflow, and the project was recognised in the national 2017 Knowledge Transfer Ireland awards, by winning the Consultancy Impact category. This successful project led to a long-term relationship with Rusal Aughinish Alumina, including the company supporting and providing co-funding for the interdisciplinary €1.9M Modelling Multiphase flow in Manufacturing grant.