AI-Enabled Predictive Analytics for Enhancing Supply Chain Resilience

Thesis Abstract

The increasing complexity and volatility of contemporary global supply chains have underscored the urgent need for innovative strategies to bolster resilience against disruptions. This study investigates the application of artificial intelligence (AI)-enabled predictive analytics as a transformative approach to enhancing supply chain resilience, addressing the critical challenge of real-time decision-making in uncertain environments. The primary aim is to develop a comprehensive framework that integrates AI-driven predictive models to forecast potential supply chain disruptions and inform proactive mitigation strategies. Specific objectives include identifying key predictive variables influencing supply chain stability, evaluating the effectiveness of machine learning algorithms in risk prediction, and proposing an operational model for implementing AI-driven analytics within supply chain management systems. Employing a mixed-methods research design, the study combines quantitative data analysis with qualitative insights to yield a holistic understanding of the subject. The target population comprises supply chain managers, logistics coordinators, and IT professionals within manufacturing firms operating in a metropolitan region with a substantial supply chain footprint, totaling approximately 150 organizations. A stratified random sampling technique was used to select 60 firms, from which 120 supply chain practitioners and 30 IT specialists were purposively sampled for in-depth interviews and surveys. Quantitative data were collected through structured questionnaires measuring variables such as predictive accuracy, decision-making turnaround time, and resilience indicators, while qualitative data were gathered via semi-structured interviews exploring contextual factors influencing AI integration. The validity and reliability of the research instruments were established through pilot testing, expert validation, and Cronbach's alpha coefficients exceeding 0.85 across all survey scales. Quantitative data analysis involved descriptive statistics to outline key characteristics, followed by multiple regression analysis to assess the relationship between predictive analytics adoption and supply chain resilience. Machine learning techniques—specifically, Random Forest and Support Vector Machines—were employed to develop and validate predictive models, with model performance evaluated using accuracy, precision, recall, and F1 scores. Qualitative data were subjected to thematic analysis using NVivo software, enabling the identification of emergent themes related to organizational readiness, technological challenges, and change management strategies. The integration of findings was facilitated through triangulation to validate model robustness and contextual relevance. Expected findings anticipate a positive correlation between the deployment of AI-enabled predictive analytics and enhanced supply chain resilience, evidenced by improved risk detection accuracy, reduced lead times in response to disruptions, and increased adaptive capacity. The study also foresees identifying critical predictive variables—such as supplier failure probabilities, inventory variability, and transportation delays—that significantly influence resilience outcomes. Furthermore, the research aims to demonstrate that machine learning models outperform traditional statistical methods in forecasting disruptions, thereby enabling more agile responses. This research contributes to the existing body of knowledge by providing a validated framework for integrating AI-driven predictive analytics into supply chain operations, grounded in the Dynamic Capabilities Theory and the Technology-Organization-Environment (TOE) framework. It advances practical understanding of how manufacturing firms can leverage cutting-edge AI techniques to foster resilience proactively, offering strategic and operational implications for supply chain stakeholders. The study concludes that organizations adopting AI-enabled predictive analytics are better equipped to anticipate disruptions, allocate resources efficiently, and maintain continuity during crises. Based on these findings, comprehensive recommendations are made for managerial practice, including investing in AI infrastructure, enhancing data quality, and fostering organizational agility. The research also proposes avenues for future inquiry, particularly exploring longitudinal assessments of AI's impact on supply chain resilience across different industrial contexts.