Integrating AI and ML in Supply Chain Digital Twins: Bridging Potential and Foundational Research Gaps

Supply chains face increasing complexity, volatility, and demands for resilience, sustainability, and real-time responsiveness. These challenging conditions highlight the limitations of traditional optimization methods. The integration of artificial intelligence (AI) and machine learning (ML) with digital twins offers transformative potential for addressing these challenges through enhanced analysis, prediction, and automation. This chapter examines the critical synergy between AI/ML and digital twins specifically for supply chain optimization. It details how AI/ML capabilities can power intelligent data processing, advanced predictive analytics, predictive maintenance of assets, and pathways toward autonomous decision-making within supply chain digital twins (SCDTs). However, moving beyond conceptual potential, the chapter critically analyzes the significant foundational research gaps that currently impede the reliable, scalable, and trustworthy deployment of these integrated technologies. Key challenges synthesized include the pervasive need for robust and continual verification, validation, and uncertainty quantification (VVUQ) methodologies suited for dynamic AI/ML-driven systems, the development of theoretically sound and practical hybrid modeling approaches, scalable integration across complex systems of systems, managing heterogeneous and imperfect supply chain data, and addressing critical ethical, governance, and cybersecurity considerations. The chapter concludes that realizing the future vision of intelligent, optimized, and resilient supply chains is fundamentally contingent upon a concerted, collaborative, and interdisciplinary research effort focused on overcoming these mathematical, statistical, computational, and translational foundations, with a paramount focus on building trust and ensuring reliability.