FoodLifeTimeTracking: Use of multimodal information fusion for the realization of a monitoring device and a life-cycle simulator for the investigation and quantification of quality-determining parameters and the expiry date of foodstuffs and their ingredients.

The prediction of the stability of food and intermediate products as best-before or use-by dates, in terms of accuracy and reliability, is to be significantly improved by the findings in this project. As a result, product safety can be increased, production costs can be reduced and food waste can be reduced. The information on basic aging effects in food products will be used to generate data-based, adaptive models that can be used to predict stability/shelf life much more accurately.

A key aspect of this is the assessment of the physicochemical state of foods and food intermediates. Relationships between product properties, environmental parameters and stability are analyzed and visualized using AI-based techniques such as machine learning and information fusion. Standardized aging procedures and forcing methods are used to identify indicators of fundamental aging effects in order to generate data-based AI models that enable reliable prediction of food stability.