Human-Centered Multi-Objective Flow Shop Scheduling with Missing Operations under an Industry 5.0 Perspective

Abstract

Industry 5.0 is transforming production planning by integrating human factors into decision-making. This paper addresses a multi-objective permutation flow shop scheduling problem with missing operations, where jobs may skip machines and processing times depend on worker-related factors such as skill level, age, learning, and forgetting. The model simultaneously minimizes makespan, total tardiness, and total earliness. To solve the problem, an NSGA-II-based evolutionary algorithm with permutation encoding was implemented. Computational experiments with 0%, 20%, and 40% missing operations show that greater route flexibility generally improves schedule quality, reducing makespan by 4.41% to 27.71% and total earliness by 18.11% to 77.53% in compromise solutions, while tardiness shows a more heterogeneous behavior. The results also indicate that missing operations reshape the trade-off structure among objectives rather than simply reducing problem size. Overall, the study provides a human-centered scheduling approach aligned with Industry 5.0 and relevant for production environments with workforce heterogeneity and route variability.