Artificial Intelligence and Predictive Analytics for Supply Chain Risk Management: Enhancing Intelligent Decision-Making

Going from reactive risk management to proactive and data-driven approaches to managing risks in the supply chain is a necessary step in today's operations management challenges. The study explores how Artificial Intelligence (AI) and predictive analytics could be applied to improve intelligent decision-making in global supply chains. The traditional approaches are often unsuitable to handle the complex and non-linear nature of the modern supply chain disruption and require advanced computational techniques to maintain operational continuity. This study examines an AI-based predictive model in comparison to traditional statistical models, using a complete empirical database from a multi-level manufacturing and logistics supply chain. Evaluation of these models is done in the following three key operating characteristics: Mean Absolute Percentage Error (MAPE) in the context of demand forecasting during periods of market volatility; Time-to-Recovery (TTR) after a severe disruption event across different tiers of the supply chain; and risk mitigation efficiency in comparison to the prediction lead time required for identification. The findings of the empirical study show that the AI-driven framework significantly cuts down the error in demand forecasts by 61.2% for a 12-month test period. Additionally, the predictive model shortens recovery operations with up to a 73.1% TTR reduction in warehousing operations. Last but not least, it is shown that there is an exponential relation between early predictive lead time and risk mitigation. The findings are testament to the strength of machine learning algorithms in dealing with high-dimensional data and demonstrate that predictive analytics has the potential to revolutionize the concept of enterprise resilience..