Natural fluctuations of the quality of raw materials and varying environmental conditions during the manufacturing process of foodstuff result in individual risks of spoilage as well as pollution by harmful substances. However, manufacturers determine a universal expiration date for a manufacturing process of a formula. Estimating a shelf life individually or by batch is not feasible with the currently applied methods in industrial practice, as they include lengthy storage tests.
Furthermore, check-ups for pollution take place in general and not on individual products, since laboratory analyses are time-consuming and costly. The project will explore innovative modelling technologies based on rapid data acquisition and data analysis in order to predict the expiration date more accurately. In the process, economical sensor technologies provide comprehensive results of every individual product’s condition.
Innovative technologies, which use information fusion and machine learning for a more accurate prediction of the expiration date stated on the food packaging, can also contribute to the reduction of food waste. The systems for real time quality control integrated in the production necessarily for that purpose represent an important contribution for food safety.