Article Preview
Top1. Introduction
The delicacy and complexity of food has resulted in the ever-increasing issue of food safety (George et al., 2019). Therefore, food safety has become a policy of concerned in many countries around the world. With globalization and the current complications of collaborative food supply chains, the complexity and uncertainty of food safety issues have increased, and the maintenance of food safety has become more difficult. Any part of the supply chain may be a cause of food hazard due to human, machinery, material, regulatory/procedural, or environmental factors.
Thus, food safety management should include the entire food supply chain in the monitoring and management scope, and consider all stakeholders and influencing factors (ISO, 2018). The stakeholders of food safety include all the operators in the collaborative food supply chain, governmental food supervision departments, international food safety organizations, and consumers. The influencing factors of food safety include the attributes of human, machinery, material, regulatory/procedural, and environmental factors in the food lifecycle. In terms of raw material supply, the complicated sources and increased pollution sources have resulted in more and more contaminated foods, or foods with illegal chemical additives (Resende-Filho & Hurley, 2012). At the same time, some suppliers in the food supply chain lack food safety management capacity, are unaware of their responsibilities, and lack integrity, thus resulting in deliberate falsification or cover up, and leading to information asymmetry between consumers and manufacturers (Sun & Wang, 2019). The above-mentioned problems may cause food safety incidents. Therefore, the question of how to effectively and comprehensively manage the food safety issues of collaborative supply chains is difficult in both method and technology.
The advancements in computer and network technologies have enable the rapid accumulation of data and information, as well as the application of big data in various industries. The significant increase in data and the upgrading of the computing speed gives rise to the effective utilization of artificial intelligence and machine learning. In recent years, some scholars have applied machine learning in the food industry for the prediction of crop yields (Chlingaryan et al., 2018; Mehra et al., 2018), life cycle evaluation (Nabavi-Pelesaraei et al., 2017), and logistics and distribution management (Krisztin, 2018). However, the abovementioned research only focused on the quality prediction and management of the various stages of the supply chain, but lacked quality management and anomaly protection within the entire supply chain.
Food safety supervision and management of the entire supply chain is necessary, and must be able to: (1) track the supply chain of food raw materials; (2) ensure the autonomous management of all supply points in a supply chain; and (3) detect anomalies in the supply chain and analyze the causes of such problems, while simultaneously track the other affected foods (Wang & Yue, 2017).
Food safety is realized upon a comprehensive food protection architecture, as well as the collection and analysis of the data related to food safety in supply chains (Hazen et al., 2014). However, the food supply chain is complicated and diversified with various types of data from each stage of the lifecycle. How to define and collect useful data related to food safety in the supply chain, in order to facilitate food safety supervision and management in the food lifecycle, is an important topic for food safety management (Kamble et al., 2020). Information sharing and coordination among organizations have become complicated in today’s globalized food chain (Morgan et al., 2018). In addition, technologies for the understanding and prediction of food safety incidents need to be further developed based on the food safety data in the supply chain, so as to prevent anomalies.