AI-Driven Resilient Supply Chain Architectures: Machine Learning Frameworks for Risk Anticipation, Disruption Mitigation, and Adaptive Decision-Making

Abstract

Global supply chains face various disruptions related to demand volatility, geopolitical tensions, weather-related disruptions, and operational failures, making conventional rule-based and reactive resilience approaches inadequate for addressing these disruptions. Recent developments in artificial intelligence (AI) and machine learning (ML) offer promising opportunities for transforming supply chain resilience from conventional reactive approaches to proactive, adaptive, and learning-enabled decision-making systems. However, the prevailing literature primarily focuses on standalone AI-related activities without discussing the architectural aspects of integrating AI for end-to-end supply chain resilience under uncertainty. This study proposes a unified resilient supply chain architecture with the incorporation of machine learning technologies in the processes of anticipation, mitigation, and adaptive decision-making. The research methodology used in the study is the design science research approach, which resulted in the development of a multi-layered framework consisting of the data intelligence layer, the risk anticipation layer, the prescriptive decision and mitigation layer, and the adaptive learning feedback layer. The framework ensures the integration of various machine learning algorithms with supervised learning, unsupervised learning, and reinforcement learning techniques with various categories of supply chain risks. In order to test the theoretical robustness of the proposed architecture, a scenario-based evaluation approach is followed, and various disruption scenarios, which cover failures from the suppliers, disruptions in the logistics network, and demand disruptions, are analyzed. New evaluation criteria, which are classified under the term "resilience-oriented evaluation," are also proposed in the paper, which move beyond the traditional evaluation criteria of costs and services. This paper contributes to the literature in the fields of supply chain and operations management, as an integrative AI architecture for building resilience is proposed, and the conceptualization of adaptive supply chains is advanced, while also providing guidance for managers and policymakers who are interested in leveraging AI as a strategic tool for building resilience in an ever-increasingly volatile environment.