Artificial intelligence based supply chain management strategy during COVID-19 situation

The COVID-19 pandemic has brought about unprecedented challenges, such as constantly changing consumer demands and disruptions in the supply chain. This study aims to reduce the cost of inventory units in the COVID-19 scenario by using a mathematical model to develop fundamental inventory control formulations. However, conventional analytical methods often struggle to identify and analyse the complex patterns and rapidly shifting trends that characterise COVID-19. To address this issue, this study employs ANN techniques to determine the total variable cost of essential items in the COVID-19 scenario, using current demand, discount rate, and advanced payments as input variables. The study creates an optimal ANN model for real-life situations by combining different solvers and activation functions. Among the various combinations, the study observes that the combination of the relu activation function and lbfgs solvers offers the lowest RMSE value of 0.03382 and the highest R2 value of 0.9858 for the optimum total variable cost. The study affirms that ANN techniques excel in analyzing the EPQ model during COVID-19 by managing complexity, adapting to changes, and assessing diverse data. This real-time analysis aids supply chains, ensuring timely key item availability in high-demand or disruptive scenarios.