Key takeaways
Short answer: Setup time is the clock on readying a single machine. Changeover is the entire product switch across the line, setup plus material change, first-good-part and ramp-up to full speed. SMED reduces both, but if you only measure machine setup you miss most of the loss; changeover is where OEE Availability and Performance really bleed on high-mix lines. See also oee for manufacturing.
Setup time is narrow and machine-centric: the clock from the last good part of one job to the machine being ready for the next. It is easy to measure and easy to feel proud of, which is exactly why it can mislead.
Changeover is the whole switch as the line experiences it. It includes every machine setup, but also the material and consumable change, the first-good-part confirmation, and the ramp from a crawling restart up to full rate. This is the number that actually costs you output.
A plant proudly reports 10-minute setups. But the full changeover tells another story: 10 minutes of machine setup, then 15 minutes waiting for the next material lot, 12 minutes producing scrap until the first good part, and 18 minutes ramping to full speed, 55 minutes of real changeover, not 10.
Optimising the 10-minute setup further is almost pointless; the money is in the other 45 minutes that "setup time" never counted.
SMED (single-minute exchange of die) converts internal steps, done while the machine is stopped, into external steps done while it still runs, then streamlines the rest. Applied to one machine it cuts setup; applied to the whole switch, staging materials in advance, standardising first-article checks, pre-loading programs, it cuts changeover, which moves OEE far more.
Measure the wrong clock and you optimise the wrong thing. A line that tracks only machine setup can shave seconds off a number that barely matters while an hour of material-handling and ramp loss goes untouched at every changeover. Define changeover end-to-end, last good part of A to stable full-rate good parts of B, and the real opportunity appears.
Changeover hits Availability while the line is stopped and Performance during the slow ramp. On high-mix lines it is one of the largest and most addressable OEE losses, which is why reducing changeover, not just setup, is such a high-leverage move.
1. Measuring setup, not changeover. Most of the loss is invisible.
2. Ignoring ramp-up. The slow restart is real Performance loss.
3. Not counting first-article scrap. Quality loss during changeover gets missed.
4. Treating SMED as a one-off event. Without standard work, gains erode.
Fabrico times changeovers end-to-end, including ramp and first-article loss, so you optimise the real number, not just machine setup. Book a demo to see full changeover loss on your lines.
No, it spans the whole line plus material change, first-good-part and ramp-up, which machine setup never counts.
Both, by converting internal steps to external; the biggest gains come from attacking the full changeover.
Changeover, measured end-to-end, it captures the Availability and Performance loss setup time hides.
Within changeover, readiness is not done until good parts flow at rate.
Because the loss is in material handling, first-article scrap and ramp, the parts "setup time" excludes.
Login
Blog
