Predictive Drug Inventory Management Using XGBoost, ARIMA, and IoT Technologies

Abstract

Drug inventory is a critical component of healthcare systems; however, it is frequently managed through manual or basic processes, leading to medicine shortages, overstock, delivery delays, and waste due to expired medications. Recent advancements emphasize machine learning and data-driven techniques for drug inventory management. Methods such as XGBoost, Random Forest, and ARIMA are utilized to enhance demand forecasting and trend analysis, while classification algorithms identify medicines at risk of shortage. Essential features, including real-time monitoring, tracking, and secure access, are recognized as vital for system reliability and efficiency. Despite the availability of numerous methods, most solutions address only isolated aspects and lack a comprehensive approach. Consequently, there is a need for an integrated system that combines prediction, monitoring, and security to improve decision-making, reduce waste, and enhance the overall effectiveness of healthcare inventory management.