Variations in milk quality are a daily challenge when producing the traditional creamy dessert Tocinillo de Cielo. When quality varies from batch to batch, manufacturers run the dual risk of product returns from dissatisfied consumers and poor utilisation of raw materials.

This is why the family-owned producer Dulcegrado has been working with AOTECH on an S3FOOD funded project to evaluate real-time characterisation of milk quality – primarily fat and protein content – using NIR spectroscopy.

Using this data, Dulcegrado is able to adjust the addition of powder milk and cream to ensure a uniform texture, flavour and firmness with every dessert batch. Apart from keeping consumers satisfied, improved knowledge about milk quality will ensure only the right amount of powder milk and cream is added to standardise the milk – securing efficient raw material use.

This project’s main result has been the successful demonstration of NIR technology potential for the characterisation of certain milk parameters, especially the percentages of fat, protein, lactose and lean dry matter. These parameters have been obtained with fairly good prediction errors, ​fulfilling the objectives of the project.

Data gathered by NIR spectroscopy will be in readily available cloud storage, eliminating the need to send milk samples for analysis at an external laboratory and supporting traceability from raw material to retail product.

Final report summary
The project has successfully demonstrated the potential of NIR technology for characterising milk parameters such as fat, protein, lactose and lean dry matter content. An additional finding is that information obtained using the AONIR-C equipment can be easily integrated into an IoT platform. In other words, the information can be stored, visualised and processed to obtain valuable data about the milk.

S3FOOD has provided funding for a longer evaluation of the technology to validate its advantages. A description of this project is available here.