A Multidimensional Framework for Evaluating Operational Learning in Human-Centered Assembly Systems: Effects of Digital Assistance and Virtual Reality

The digital transformation of manufacturing requires training approaches that build operational competence in human-centered production systems. This study develops and experimentally applies a multidimensional framework to characterize operational learning and compare alternative operator training strategies. Experiments were conducted in a Lean Learning Factory using three strategies, conventional workstation training, digitally assisted workstation training, and virtual reality-based training, followed by practical validation. Operational learning was assessed across five dimensions—accuracy, continuity, efficiency, convergence, and retention—using statistical analysis, regression modeling, convergence indicators, and maturity-stage assessment. The results show that the three strategies yield distinct learning trajectories. Digital assistance was associated with faster development and stabilization of operational performance, whereas virtual reality training was associated with higher initial performance during subsequent practical validation. The proposed OLM–OLAM–OLMM framework integrates these dimensions into a common methodology for describing, comparing, and assessing the maturity of operational learning. The approach extends traditional learning-curve analysis by incorporating performance convergence, retention, and maturity assessment, providing a structured basis for evaluating operator training strategies in human-centered Industry 5.0 manufacturing systems.