Efficient logistics and transportation systems are essential for reducing post-harvest losses and enhancing supply chain performance in agriculture. This research introduces an IoT-enabled Gradient Boosting Model (GBM) designed to enhance agricultural logistics by integrating real-time sensor data and predictive analytics. Internet of Things (IoT) equipment, such as GPS modules, temperature and humidity sensors, and load cells, was installed on transport trucks and storage units to gather continuous data over a 6-month field experiment, including 120 delivery routes. GBM was trained on 15,000 data sets and achieved a 94.2% accuracy in forecasting delivery delays, improving Random Forest (91.6%) and SVM (88.7%). The model optimized route choices, achieving a 17.3% reduction in fuel usage and a 22.8% decrease in average delivery time. Moreover, cold-chain integrity was preserved 96.5% of the time, compared to 87.9% in the baseline system. The findings indicate that the integration of IoT with Gradient Boosting markedly improves decision-making in agricultural logistics, yielding enhanced efficiency, diminished operating costs, and superior post-harvest quality preservation. This method facilitates scalable and data-informed management of agricultural transportation networks. © 2025 IEEE.