ML-Based Shelf Life Prediction in Food Storage using Kinetic Models

Efficient warehouse management, supported by Wireless Sensor Network (WSN) technologies, plays a crucial role in the preservation of perishable goods. An "intelligent" storage organization, capable of monitoring and controlling environmental conditions such as temperature and humidity, can significantly improve the average lifespan of products, minimizing the risk of spoilage and waste. This paper presents a Machine Learning-based (ML) shelf-life prediction algorithm that can be integrated in the WSN sensor-based platform for active shelf-life management of stored food products, which leverages highprecision temperature and humidity sensors to acquire entrusted environmental data, to feed a ML and kinetic-based algorithm for accurate shelf-life prediction. Demonstrated through a case study on a sample batch of yellow potatoes, this adaptable model can be customized for different perishable goods by suitably selecting quality indices, thereby optimizing quality of the perishable goods and significantly reducing waste.