Development of a RFID-cloud-based location assignment and tracking system for the packaged food industry

Abstract

Food safety is increasingly of concern to the public. To ensure the quality of packaged food, storage operations along the supply chain play critical roles. In addition, the trend of e-retailing has posed additional pressure on the warehouse, therefore the warehouse has to be more efficient. Among the various storage operations, storage location assignment is a complex yet important decision to make. Assigning a suitable storage location for food can prevent food deterioration caused by the reactivity between food, food packaging and the physical environment. Besides, suitably locating products can improve the operational efficiency of the facility. On the other hand, having an effective tracking system can quickly help to retrieve any problematic food lots from the supply chain, which in turn can reduce the number of potential victims of such food. Thus, a packaged food warehouse should be equipped with a system for both assigning suitable locations for various food products, and in facilitating the external and internal tracking of the products. However, the existing decision support system (DSS) for storage location assignment seldom considers the storage needs of the food products, and the existing Radio-Frequency-Identification-based tracking system has rarely been integrated with a DSS to tackle integrative storage location assignment and product tracking. Therefore, this research proposes a comprehensive DSS called the RFID-Cloud-based Location Assignment and Tracking System (R-CLATS) for the packaged food industry. The system consists of four modules: The Data Capturing Module, Information Consolidation Module, FARM Variable Selection Module and Location Assignment Module. The Data Capturing Module applies Radio Frequency Identification (RFID) technology to collect real-time data regarding the inbound and outbound activities in the warehouse, in order to track product locations. The Information Consolidation Module uses a cloud-based database to consolidate data and information from the warehouse and external shippers in real time for further processing. The FARM Variable Selection Module applies FARM to identify the most relevant input variables regarding the storage time of products and predicts the range of storage time of products. Finally, the Location Assignment Module employs both fuzzy logic and association rule mining to assign locations for the packaged food, according to the product and package characteristics and the product's inter-relationships. The developed system was validated through a case study of a packaged food wholesaler, who runs a packaged food warehouse in Hong Kong. The pilot run of the system in the case company was successful, which means the system design is feasible. The results of R-CLATS implementation was then used to evaluate the performance of the system. The evaluation results proved that after implementing the system, the operational efficiency of the warehouse was enhanced, the quality of products can be maintained and product traceability was enabled in the warehouse. In addition, recommendations given by this fuzzy-based approach were proven to be more precise than those provided by a rigid rule-based approach. Therefore, R-CLATS has been shown to be a feasible and effective DSS for enhancing the overall performance of a packaged food warehouse.