Artificial Intelligence-Driven Supply Chain Resilience A Decision-Making Framework for Sustainable Operations and Performance Management

Abstract

Environmental uncertainty and impacts, transportation challenges, and operational inefficiencies have made it more necessary than ever to rely on intelligent decision-support systems to enhance the resilience of the supply chain, while maintaining sustainability. The study suggests an AI-based decision-making model to predict supply chain resilience and aid sustainable operational performance. A quantitative approach that leverages simulation modeling was used and involved a dataset of 150 observations of 30 operational scenarios in five SKUs, three warehouses, and three geographical regions. The Composite Supply Chain Resilience Index was created based on operational, environmental, logistics and sustainability components. Five supervised machine learning algorithms were tested with an 80:20 split between training and testing sets, and five-fold cross-validation. Descriptive statistics, correlation analysis, feature importance analysis, and sensitivity assessment were carried out to find the key factors driving resilience. The results showed that XGBoost model had the highest predictive performance with a score of (R2) value 0.8536, which was better than the other models. The most influential factors for resilience were transportation distance, lead time, carbon emission factor, flood severity, and rainfall, and the most negative relationship with the resilience index was found in transportation emissions. The proposed framework effectively combines predictive analytics, sustainability assessment, and operational optimisation in a comprehensive decision-support system. It not only offers practical guidelines for enhancing logistics planning, disruption management, and sustainable supply chain operations but also lays a strong methodological groundwork for future studies on logistics resilience using AI.