Changeover Time Calculator

External time represents work that can be done while the machine keeps running the previous job, so by definition it doesn't add to the machine-down window that internal time occupies. The improved changeover figure and every dollar savings figure derived from it depend only on internal time — moving more work into the external category is exactly the improvement SMED targets, but the value of that move shows up as a reduction in internal time, not as a separate line item.

It depends heavily on your specific process, but classic SMED case studies commonly find that a meaningful share of tasks currently treated as internal — because nobody organized the workflow otherwise — can be moved to run while the machine is still producing, through steps like pre-staging tooling, pre-heating components, or using standardized quick-connect fixtures. The improved changeover figure here assumes your entered internal time already reflects genuinely internal-only work; if you haven't yet separated tasks this way, the real achievable savings could be even larger than what this calculator shows.

Machine downtime has a real, fully-loaded cost — lost production capacity, idle labor, and overhead all continue accruing during a changeover — so converting minutes saved into an annual dollar figure translates an operational improvement into the language that actually justifies investing time and resources into a SMED project. A changeover reduction that looks modest in minutes can represent a substantial annual cost once multiplied across dozens of changeovers a week and a machine's fully-loaded hourly cost.

Many SMED practitioners aim for single-digit-minute changeovers as an aspirational target, achieved through a combination of moving tasks to external time, eliminating unnecessary steps entirely, and parallelizing remaining internal work across multiple people. Where you land realistically depends on your specific equipment and process — the improved changeover figure here reflects only converting internal time you've already accurately separated from external, not further elimination or parallelization opportunities beyond that first step.