Enhanced Food Safety Through Deep Learning for Food Recalls Prediction

Contributing authors

Abstract

Several application domains/sectors such as logistics, healthcare, industry and transportation, are exploiting the added value of deployed sensors to obtain information relevant to the domain and exploit it in different contexts (e.g. for processes optimization, for actions adaptation, for decision support, etc.). The same applies to the agriculture sector, through the deployment of smart devices and sensors that provide a wealth of datasets for irrigation tuning, crops assessment, food supply chain operations monitoring, etc. Furthermore, emerging machine and deep learning data analytics techniques are utilized as a means to obtain insights and optimize the aforementioned processes. In this context, one significant challenge refers to the enhancement of the food safety across the food supply chain given that goods and products can become unsafe for plenty of reasons, such as mislabeling allergens, contamination etc. To address this challenge, in this paper we introduce a set of deep and machine learning techniques employing time series forecasting to provide insights regarding the risk associated with each product category concerning potential food recalls. Additionally, we propose an approach based on reinforcement learning which utilizes historical recall announcements for predicting future recalls (by their type) that leads to timely recalls and contributes to enhanced food safety across the supply chain. We also evaluate and demonstrate the effectiveness and added-value of the proposed approaches through a real-world scenario that yields promising results.