A Comprehensive Review of IoT-Based Smart Pharmacies for Optimizing Stock Management with Siamese Heterogeneous Convolutional Neural Networks

Abstract

The rapid advancement of digital technologies has significantly transformed the healthcare and pharmaceutical industries, leading to the development of intelligent systems such as IoT-enabled smart pharmacies. These systems utilize interconnected sensors, cloud platforms, and data analytics tools to enable real-time monitoring of drug inventories, automated dispensing, and efficient supply chain coordination. By continuously tracking stock levels and environmental conditions, IoT-based pharmaceutical systems enhance operational efficiency, reduce medication shortages, and improve decision-making in inventory management. Alongside IoT, artificial intelligence and deep learning techniques play a crucial role in analyzing large volumes of pharmacy data generated from connected devices and management systems. Machine learning models help predict drug demand, detect anomalies in stock patterns, and recommend optimal restocking strategies to maintain balanced inventory levels. Among advanced approaches, convolutional neural networks demonstrate strong capability in extracting meaningful patterns from complex pharmaceutical datasets. More recently, Siamese heterogeneous convolutional neural networks have emerged as a powerful architecture for integrating multiple data sources such as prescription records, patient demand behavior, and supply chain information. By learning relationships between heterogeneous inputs, these models improve predictive accuracy and support intelligent recommendation systems. This review examines IoT-based smart pharmacy systems with a focus on deep learning techniques, IoT architectures, and inventory optimization frameworks, highlighting current trends, challenges, and future research directions in intelligent pharmaceutical management systems.