The aim of this study was to develop and functionally evaluate a portable, reconfigurable workstation for controlled rapid-changeover simulation based on the Single-Minute Exchange of Die (SMED) approach. The device was designed to reproduce selected structural and organisational features of a machine changeover and to allow users to identify losses, introduce non-permanent modifications and measure their combined effect. The evaluation involved 30 four-person teams. Each team carried out two consecutive changeovers of the same workstation: the first in the baseline configuration and the second after implementing SMED-oriented improvements. Internal changeover time was used as the main performance measure, and the modifications introduced by the teams were grouped according to SMED principles. Mean internal changeover time decreased from 31 min 22 s to 7 min 32 s, corresponding to a mean saving of 23 min 50 s and a reduction of 75.32%. The difference was statistically significant, with a very large effect size, and internal changeover time decreased in all teams. Because the second changeover always followed the baseline trial, the observed reduction should be interpreted as the combined effect of the implemented modifications and repeated task performance, including increased familiarity with the workstation and task. After the improvements, 23 of the 30 teams achieved an internal changeover time below 10 min. The implemented modifications included external preparation of selected activities, simplified fastening and sheet attachment, standardisation of screws and tools, visual positioning aids and elimination of redundant nuts. The results show that the workstation provides controlled and standardised conditions for reproducing changeover losses, implementing alternative improvement paths and quantitatively evaluating the combined effect of SMED-oriented modifications.