<![CDATA[Modern high-mix, low-volume manufacturing faces significant downtime during setup changes in part-feeding systems, yet no quantitative model currently exists that links group-based reconfiguration strategies to a measurable flexibility index under stochastic batch-size conditions. This study therefore aims to develop and experimentally validate a probabilistic mathematical model for assessing machine flexibility when a group-based reconfiguration approach is applied to part-feeding systems. The methodology combines Monte Carlo simulation to model random batch-size distributions with physical validation using a rotary orienting device across eight distinct sleeve types. Simulation results indicate that the proposed strategy reduces setup labor by 51-61% in systems handling 100 different part types. When fewer than one-third of parts require reconfiguration, the machine flexibility index reaches 0.088 ± 0.014, meeting established thresholds for high system flexibility. Experimental tests confirm that a uniform group-level adjustment maintains operational efficiency deviations within 3-5% across varying part geometries. The primary novelty of this work lies in introducing a confidence-bounded flexibility coefficient that explicitly incorporates auxiliary loading subsystems, which are consistently overlooked in existing deterministic approaches. This provides a practical, data-driven tool for production planning that enhances responsiveness without sacrificing throughput or increasing system complexity.]]>
