<![CDATA[Rapid and non-destructive monitoring of rice quality during storage is essential for supporting effective warehouse management. This study aims to develop and evaluate the integration of Portable Near-Infrared Spectroscopy (NIRS) with an Internet of Things (IoT) framework for real-time rice quality monitoring. A quantitative experimental approach was employed by acquiring NIRS spectra in the wavelength range of 740–1070 nm from fresh and aged rice samples. The spectral data were automatically transmitted via the IoT system to a centralized server for storage and analysis. Rice quality parameters, including moisture, fat, and protein content, were predicted using a Partial Least Squares Regression (PLS-R) model based on raw spectra without spectral pretreatment. The results indicate that the PLS-R model achieved good predictive performance for moisture and fat content, with validation correlation coefficients (R) ranging from 0.87 to 1.00 and Residual Predictive Deviation (RPD) values of 1.11–3.65 for moisture and 3.70–4.65 for fat in both fresh and aged rice samples. In contrast, protein prediction showed limited accuracy, particularly in fresh rice samples with an RPD value of 1.79. The IoT system primarily functioned as a real-time data acquisition and transmission platform, enabling integrated rice quality monitoring. Overall, the findings confirm that NIRS–IoT integration is feasible for monitoring rice quality based on moisture and fat content during storage.]]>
