Artificial Intelligence in Supply Chain Optimization: A Systematic Review of Machine Learning Models, Methods, and Applications

Abstract

Modern supply chains face immense complexity and uncertainty, driving the powerful integration of Artificial Intelligence (AI) and Machine Learning (ML) for data-driven prediction and mathematical optimization to enable robust decisionmaking. To understand the state of this critical intersection, this paper presents a systematic literature review analyzing 199 articles focused on Tangible Supply Chains. Our analysis also categorizes how ML is used, with most applications falling into either predicting model parameters or directly generating solutions. This paper introduces a comprehensive taxonomy spanning ML roles, problem types, and optimization models, and synthesizes trends, such as the dominance of Reinforcement Learning in logistics-focused formulations. Beyond providing a detailed classification of the field, this review highlights critical research gaps and contributes a novel research framework designed to guide researchers. This work serves as an essential resource for understanding current trends and identifying future opportunities at the confluence of ML and supply chain optimization.