Deep-Learning-Driven Dynamic Demand Forecasting for Emergency Medical Supplies in AIoT-Enabled Green Supply Chain Systems

The growing concern regarding the environment and the compounded nature of healthcare-related emergencies has impacted the effectiveness of medical supply chain management. As a rule, demand forecasting techniques do not synchronize the altitude of medical need with the level of environmental care throughout a public health crisis. This study presents MedSC, a deep learning framework for dynamically forecasting emergency medical supply requirements within an Artificial Intelligence of Things (AIoT) green supply chain environment. The framework features advanced fuzzy clustering with long short-term memory prediction models in medical supply prioritization using AIoT sensor data and integrating metrics of environmental care. MedSC determines the relevance and importance of medical supply through eco-friendly hierarchical clustering for prioritized grouping. It then devises unique forecasting models for noninterconnected and interconnected supply chain subsystems. The results of experiments conducted at three major medical distribution centers validate the claims in the hypothesis, proving that the framework outperforms traditional forecasting methods and even sophisticated contemporary approaches. These performed exceedingly well on ultralow-latency real-time predictive bounds, claiming an industry record in fast predictive bounds, holding double the precision while throttling emissions by 28.5%. The outstanding attribute of the framework is the balance of healthcare needs and environmental care, which improved energy per unit work by 91.3% while 89.7% improved resource per unit work.