Intelligent Evaporative Cooling Systems for Post-Harvest Fruit and Vegetable Preservation: A Systematic Literature Review

Fruits and vegetables that suffer post-harvest losses are a severe bottleneck in food systems all over the world, and it has been estimated that 30-50% of perishable fruits and vegetables are lost between farm and consumer, especially in low-and-middle-income countries (LMICs) that do not have access to reliable cold chain infrastructure. Evaporative cooling (EC) has long been known as a low-cost, energy-efficient alternative to mechanical refrigeration, particularly in hot, dry environments. The conventional EC systems, however, have the disadvantage of being statically operated, climate-dependent, and inefficiently controlled, thereby restricting their utilisation and utility. A combination of the Internet of Things (IoT), machine learning (ML), and advanced control theory in recent history has led to the emergence of intelligent evaporative cooling systems (IECS)-adaptive, data-driven platforms through which real-time environmental regulation and predictive maintenance can be undertaken and autonomous optimisation achieved to obtain an improved preservation of the post-harvest of goods. The review indicates that the temperatures of the refrigerators can be reduced by 8 to 15°C, extend shelf life by 50-100%, and the energy used can be cut by 75-90% compared to traditional refrigeration. Economic analysis indicates payback periods as short as 1.2 years and total system costs of less than USD 100 for IoT components. There is a strong, consistent (R2) relationship of 0.98 or higher between machine learning models, especially Long Short-Term Memory (LSTM) networks and tree-based ensembles, and microclimate variables. Despite these improvements, critical research gaps that require attention include limited validation in tropical and high-humidity settings, the lack of standardised Ag-IoT protocols, the lack of life-cycle and Food-Energy-Water (FEW) nexus evaluations, and explainable AI (XAI) to promote farmer confidence. To address these issues, a new, combined 4-layer framework comprising the Physical, Sensing/Actuation, Data/Communication, and Intelligence/Control layers is proposed as the roadmap for next-generation IECS. Such an SLR concludes with a future-oriented research agenda that focuses on robustness, interoperability, sustainability, and human-centred design, and provides practical recommendations for researchers, engineers, policymakers, and agricultural practitioners who are determined to create climate-resilient and equitable food systems.