Case study of the monitoring of ultra-high temperature processed partly skimmed milk production batches by means of an electronic nose

Smell and flavor are key factors that determine the acceptance of food by consumers. With reference to milk production, some production batches, despite passing routine microbiological and physical-chemical standards, exhibit bad odors for many reasons related to the production process itself (from fresh milk delivery to bottling). In this article, we present a case study of the capability of an electronic nose in discriminating different qualities of milk by sensing the aroma of milk samples belonging to different production batches and to different brands. All the analyzed milk samples were ultra-high temperature (UHT) processed, partly skimmed, and commercially available at retailers. Milk samples showing off-flavors were compared with other samples of the same kind of milk belonging to the same brand (but to different production batches) and to different brands. The comparison was performed by comparing the smell of the samples just after opening the packaging and again 2 h later. In all cases, principal component analysis was able to discriminate samples into two different groups (normal and anomalous). Moreover, analysis of the olfactory fingerprints showed that 2 h after the opening of the packaging, the flavor of anomalous samples evolved in a different way from that of the normal samples. The results confirm the utility of the e-nose approach in monitoring the quality of UHT partly skimmed milk production batches, especially if combined with chemical, physical, and sensory technique