Predictive modelling of dairy manufacturing processes

Predictive modelling methodologies have been used to predict and control many dairy production processes, from on-farm livestock health and production processes to post-farmgate dairy manufacturing processes on an industrial scale. Mathematical modelling of end-product attributes to input variables such as raw materials or process settings provides a very powerful tool to study the complex interactions of raw materials, process settings and attributes of dairy products such as composition, functionality, sensory attributes and shelf life. Predictive modelling of many processes through the entire dairy supply chain is providing significant advantages in increased process efficiency and increased quality control, resulting in associated economic benefits to dairy producers. From a research point of view, the use of predictive modelling methodologies such as statistical modelling, artificial neural networks and fuzzy systems/genetic algorithms to relate target responses back to input factor settings is providing very efficient ways to study the complexities of interactions in dairy products.