Drying behavior of organic apples and carrots by using k-means unsupervised learning

Summary

Drying prevents food spoilage and decay through moisture removal due to simultaneous heat and mass transfer from food, which may be stored for long period with minimal deterioration occurring. However, drying technology is not always paired with good/excellent organoleptic, nutritional and/or functional properties of food. In fact, during drying the heat-sensitive substances are often destroyed and degradation processes may be exacerbated due to various and concurrent reaction mechanisms. Based on authors’ best knowledge, drying degradation kinetics of biological materials are usually pseudo first-order or first order reactions (i.e. carotenoids degradation in carrots) and may be affected by the initial quality of the product itself. Therefore, the main objective of the proposed study was to investigate the feasibility of k-means unsupervised learning to proactively monitor quality change in organic apples and carrots during hot-air drying. Based on authors’ best knowledge, fruit and vegetables drying has been widely addressed in literature; nevertheless, little insight is available on smart drying, while knowledge of its potential use in the organic sector is totally lacking.