Characteristics of Production Scheduling Problems in the Era of Industry 4.0 – A Review of Machine Learning Algorithms for Production Scheduling

Changes in customer demand and technological advances increase production scheduling complexity. A solution to handle the increased complexity is to facilitate machine learning. Current reviews of the research of scheduling algorithms focus on summarizing existing methods. They lack the characteristics of practical implications arising from the increased complexity of the scheduling problem by Industry 4.0 and do not consider how the scheduling algorithms cope with the added complexity. Therefore, we address these issues and synthesize the current literature in a comprehensive literature review and taxonomy. We discuss how they address changeovers, buffers before and after machines, prioritization of jobs, dynamics of production, and transportation on the shop floor. Afterward, we summarize the concepts of scheduling approaches and derive five research gaps.