Enhancing Post-Harvest Management Through ML and AI-Driven Spoilage Prediction and Smart Storage Solutions

Abstract -The aim of this study is to demonstrate applications of Machine Learning (ML) and Artificial Intelligence (AI) in post-harvest management in order to mitigate spoilage, provide storage condition recommendations, and ultimately achieve food security. This study presents an ML and AI approach for improving decision making in post-harvest, using predictive models based on conditions of temperature, humidity, and crop type. Random Forest and Linear Regression algorithms are then employed to classify spoilage risk levels, and recommend optimal conditions, approaching real-time conditions for practitioners. Through the combination of smart storage alternatives and AI, it is anticipated that farmers will mitigate spoiling and losses of between 20% and 30% when applied in post-harvest logistics while packaging and exploring efficiencies and shelf life in the overall supply chain enhancement. The study reports the potential of ML and AI to improve sustainability and profit and reduce loss in agriculture in post-harvest food systems.