Explainable AI and Digital Twin Integration for Agile Supply Chain Resilience in U.S. Manufacturing
Michael Ogundipe
*
Northeastern University, Roux Institute, College of Professional Studies, Portland, ME, USA.
Rhoda Kalu Tasie
Department of Industrial Engineering, University of Arkansas, Fayetteville, Arkansas, USA.
Isaac Aboh
College of Business, The Ohio State University, Columbus Ohio, USA.
Goodness Nwabueze
Micron Technology, Folsom, CA, USA.
Samuel Yaw Larbi
Northeastern University, Roux Institute, College of Professional Studies, Portland, ME, USA.
*Author to whom correspondence should be addressed.
Abstract
Global supply chains face increasing complexity due to geopolitical conflicts, pandemics, and rapid technological transformation, creating an urgent need for manufacturing systems that are both resilient and agile. This study investigates the potential of integrating Explainable Artificial Intelligence (XAI) and Digital Twin (DT) technologies to enhance agility, transparency, and decision-making within U.S. manufacturing supply chains. Although XAI and DT have demonstrated significant independent applications, their combined deployment with explainability embedded as a core design feature remains underexplored, representing a critical gap in current knowledge.
A systematic literature review was conducted, identifying 1,284 articles through database searches. After removing 418 duplicates and applying predefined inclusion and exclusion criteria, 67 articles were selected for detailed synthesis. The analysis identified five thematic dimensions central to supply chain resilience: the use of XAI in risk-sensitive decision-making, transparency across risk domains, trust and human–AI collaboration, real-time adaptation enabled by Digital Twins, and practical challenges associated with real-world implementation.
The synthesis yields a conceptual framework that integrates XAI at critical decision nodes within Digital Twin environments to address the “black-box” limitations of AI-driven systems. This framework demonstrates the potential to enhance transparency and operational responsiveness, with reported improvements including up to an 18% reduction in lead-time variability during supply chain disruptions. Overall, the findings suggest that the integration of XAI and DT technologies can significantly improve supply chain agility, stakeholder trust, and resilience by combining predictive analytics with interpretable insights.
Despite these benefits, adoption remains constrained by persistent barriers, including data fragmentation, organizational resistance, uneven sectoral readiness, and evolving regulatory requirements. This study contributes a conceptual implementation model for XAI–DT integration in U.S. manufacturing supply chains, providing a structured pathway for enhancing operational resilience while identifying key priorities for future research and industry deployment.
Keywords: Explainable Artificial Intelligence (XAI), Digital Twins (DTS), agile supply chain resilience, manufacturing industries