Abstract
<title>Abstract</title> <p>This study systematically reviews the application of machine learning (ML) models to enhance post-harvest (P-H) management and reduce losses in perishable fruit supply chains. A comprehensive literature survey was conducted across multiple databases, including Google Scholar, Scopus, IEEE Xplore, ScienceDirect, SciSpace, Web of Science, and PubMed, covering 122 studies published between 2015 and 2025. The review examined ML applications across all stages of the fruit supply chain, including harvesting, sorting and grading, storage, transportation, packaging, processing, and retail, highlighting methodological trends, performance metrics, and technological integrations. A bibliometric analysis was also conducted to identify publication trends, research clusters, and leading journals in ML applications for post-harvest management. Ensemble ML approaches, such as Random Forest, Gradient Boosting, XGBoost, and hybrid convolutional neural network–based models, consistently outperformed single ML models and traditional statistical methods in classification, prediction, and quality assessment tasks. These models support critical applications, including ripeness assessment, early defect detection, shelf-life estimation, and environmental monitoring, enabling accurate and timely data-driven decisions. Integration of ML with non-destructive sensing technologies (e.g., RGB, hyperspectral, NIR, and e-nose systems) further enhances monitoring accuracy and operational efficiency. Despite these benefits, challenges persist, including limited dataset availability, computational constraints, inconsistent evaluation practices, and restricted real-world deployment. Future adoption will require larger, more diverse datasets, standardized validation frameworks, and implementation under practical operational conditions. Overall, this review provides a unique synthesis of ML applications across all post-harvest stages, offering actionable insights to reduce losses, improve fruit quality, and advance sustainable supply chain management.</p>