Real- Time Quality Monitoring and Prediction System for Logistics 4.0 of Mango Agroindustry

Abstract

The challenge of distributing climacteric fruit is quality assurance due to the long-distance and perishability nature of the fruit. While monitoring transportation conditions is common, little research has developed a prediction model of fruit quality affected by transportation conditions. The presented study designs a quality monitoring system for mango's long-distance supply chain by integrating the Internet of Things (IoT) and machine learning. The system modeling utilizes Business Process Model and Notation and a Use Case Diagram based on requirement analysis. The design of IoT architecture addresses the needs of the supply chain actors to monitor the transportation process and predict the final quality of mango upon arrival. Artificial Neural Network (ANN) predicts mango grade classification upon arrival. The dataset consists of initial (harvest) maturity level and transportation conditions as predictor variables and mango final grade as the target variable. The accuracy of the prediction model reaches more than 95%. The verification and validation of the system with traceability technique on the user's requirements confirm the fulfillment of each requirement's input, tasks, and output. This conceptual design presents IoT and machine learning as promising solutions to quality assurance problems in the global fresh produce supply chain.