Abstract

The logistics sector is evolving rapidly with the combination of Internet of Things technologies, yet efficient decision-making in supply chain environments remained a critical challenge. The study addressed the issue of real-time visibility, delayed decision response, and inconsistent predictive accuracy in logistics operations. The background of the study was grounded in IoT-enabling tracking systems that continuously generated heterogeneous data across supply chain nodes. However, the raw data streams lacked robust analytical interpretation for actionable intelligence. The problem focused on inefficiencies in predictive logistics routing and inventory forecasting within dynamic supply chain networks. Conventional machine learning models often is limiting adaptability to evolving transportation patterns and demand fluctuations. To overcome these limitations, an Adaptive Gradient-Enhanced Ensemble Learning (AGEL) method is introduced. The proposed method is combined ensemble learning principles with adaptive weight optimization for improving prediction stability. The methodology is utilized IoT sensor data combination, feature normalization, and AGEL-based classification for demand prediction and route optimization. The system is evaluated using standard logistics performance metrics such as delivery time, prediction accuracy, and resource utilization efficiency. The results are showing that the proposed framework is improving prediction accuracy and reduced operational latency that is compared to baseline models. The proposed method is combined ensemble learning with adaptive gradient-based weight optimization. The system is achieving 93% prediction accuracy, 26% delivery time reduction, 91% resource utilization efficiency, 90% route optimization efficiency, and 85 ms system latency. The results is confirming that adaptive learning is significantly improving the logistics performance that is compared to baseline models.

Authors

A. Vinod Kanna¹, S. Muniyappan²
Sri Shakthi Institute of Engineering and Technology, India¹, Annapoorana Engineering College, India²

Keywords

Logistics IoT, Supply Chain Optimization, Adaptive Machine Learning, Ensemble Learning, Predictive Analytics