Design of Smart Cold Chain Application Framework Based on Hadoop and Spark

Abstract

A smart cold chain management is an information system which monitors and maintains the proper temperature of the products during whole cold chain, and analyzes data to report abnormal environmental status and risks to producers and consumers. In this paper, we propose an application framework for smart cold chain management system based on Hadoop, Spark and IoT (Internet of Things) techniques. The proposed application framework for smart cold chain provides PaaS (Platform as a Service) and IaaS (Infra as a Service) so as that smart cold chain management systems can be developed and operated with low cost and in short time. Also, the proposed application framework for smart cold chain allows the heterogeneous IOT devices such as RFID tags, WSN sensor nodes, BLE (Bluetooth Low Energy) sensor nodes and so on. We design PaaS and IaaS based on Hadoop and Spark to store the large amount of data stream on salable storage and process stream data in real time to detect events and assess risks in cold chain. Through generalizing the functions of existing cold chain management systems, common components in smart cold chain management systems are drawn.

Mining Data Streams for the Analysis of Parameter Fluctuations in IoT-Aided Fruit Cold-Chain

“…The system could be based on open-source technologies for machine learning and for Big Data analysis of a third generation which enable real time data processing [29]. Systems suitable for cold chain data processing are Hadoop and Spark [30]. The proposed IoT cold chain model is given in Figure 2.…”