: Modern supply chains face disruptions of growing frequency and scale, from pandemics and geopolitical tensions to natural disasters and demand volatility. Traditional risk-management methods, grounded in historical patterns and static planning, struggle to anticipate such complex events. This study examines how artificial intelligence (AI), and predictive analytics in particular, can strengthen supply chain resilience. Three machine-learning models — Random Forest regression, Gradient Boosting, and Long Short-Term Memory (LSTM) time-series neural networks — were applied to historical and case-based datasets across manufacturing, logistics, and retail sectors, and were further evaluated through scenario-based simulations of supplier delays, transportation failures, and demand spikes. The results show measurable gains: a 28 percentage-point improvement in demand forecasting accuracy (from 67% to 95%), an 18% reduction in average lead time (from 12.5 to 10.3 days), and a 17 percentage-point gain in service level under disruption (from 78% to 95%). The findings suggest that AI-enabled predictive systems help organizations move from reactive to proactive operations, reduce lead times, and sustain service levels under uncertainty. The study highlights the strategic value of integrating AI into supply chain planning to improve flexibility, responsiveness, and competitiveness.
Yadav et al. (Fri,) studied this question.