Dual-resource constrained flexible job shop scheduling with grey processing time and worker collaborative learning effect

Purpose This study addresses a dual-resource constrained flexible job shop scheduling problem (DRCFJSP) in the context of Industry 5.0, where both machines and workers must be coordinated. It incorporates a team-based collaborative learning effect that dynamically enhances worker proficiencies and uses interval grey numbers to represent uncertain processing times. The aim is to optimize makespan, maximum team workload, and total worker proficiency improvement simultaneously. Design/methodology/approach A multi-objective grey dual-resource constrained flexible job shop model is proposed, integrating interval grey numbers to represent uncertainties and a collaborative learning effect to reflect skill evolution. To solve this problem, a knowledge-guided multi-objective evolutionary algorithm (KGMOEA) is designed, featuring a four-layer encoding scheme and domain-specific strategies such as knowledge-guided population initialization and neighborhood search. Findings The proposed algorithm outperforms benchmark methods in convergence and diversity. Incorporating collaborative learning effects reduces total completion time and improves resource allocation efficiency. Experimental results confirm that the method effectively balances production efficiency with workforce skill development under uncertainty. Practical implications This paper provides a practical scheduling framework for high-complexity manufacturing environments, such as aviation composite workshops, where human–machine collaboration and skill development are critical. The proposed approach helps to optimize both production efficiency and worker proficiency growth under uncertain conditions, supporting resilient and sustainable manufacturing in the industry 5.0 era. Originality/value This paper is the first work to integrate collaborative learning effects and interval grey numbers into the DRCFJSP, addressing both dynamic skill evolution and production uncertainties simultaneously. The work contributes to the shift from machine-centered to human–machine collaborative scheduling and provides a new model and algorithm for learning-oriented scheduling in uncertain environments.