REAL TIME SHELF LIFE PREDICTION OF FRESH PRODUCE USING CNN BASED TEMPERATURE ANALYTICS

Abstract

India is one of the largest producers of fruits and vegetables globally, contributing about 14% of the world’s total production. However, a significant portion of this produce—approximately 30–40%—is lost due to inefficient storage and transportation systems, resulting in an economic loss of ₹92,651 crores. The objective of this project is to develop a deep learning-based model that uses temperature simulation data to accurately predict the shelf life of fresh fruits and vegetables, thereby enabling optimized transport and storage conditions. The proposed system employs deep Convolutional Neural Networks (CNNs) to predict how long produce will remain viable based on temperature data collected during storage and transit. This approach supports decision-making aimed at reducing spoilage and enhancing logistics. Traditionally, shelf-life prediction relied on static temperature guidelines, manual inspections, and general estimations based on historical data—methods that often led to inaccuracies and higher wastage. These conventional systems lack the precision and adaptability needed for efficient management, resulting in significant post-harvest losses. Reducing these losses is critical to ensuring food security, especially in a country like India. Machine learning models offer the potential to make precise shelf-life predictions by incorporating real-time environmental data, thereby minimizing waste, improving economic returns, and ensuring fresher produce reaches consumers. A deep learning model trained on temperature patterns can provide real-time shelf-life estimates, allowing stakeholders to make informed decisions regarding storage adjustments and route optimization. Additionally, the AI system can generate alerts for optimal consumption periods and transportation planning. This solution stands to significantly enhance efficiency and reduce the costs associated with spoilage